16:00:30 #startmeeting containers 16:00:31 Meeting started Tue Jun 28 16:00:30 2016 UTC and is due to finish in 60 minutes. The chair is hongbin. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:32 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:35 The meeting name has been set to 'containers' 16:00:37 #link https://wiki.openstack.org/wiki/Meetings/Containers#Agenda_for_2016-06-28_1600_UTC Today's agenda 16:00:42 #topic Roll Call 16:00:44 Spyros Trigazis 16:00:46 murali allada 16:00:51 Madhuri Kumari 16:01:00 Ton Ngo 16:01:02 o/ 16:01:03 Adrian Otto 16:01:23 tonanhngo: you changed your IRC nick? 16:02:06 Thanks for joining the meeting strigazi muralia_ mkrai_ tonanhngo dane_leblanc adrian_otto 16:02:11 The nickserver made me, I will see if I can get back to tango :) 16:02:17 #topic Announcements 16:02:28 I have no announcement 16:02:37 Any announcement from our team member? 16:03:06 #topic Review Action Items 16:03:12 1. muralia confirm the date and location for the Magnum midcycle (DONE) 16:03:18 Date: Aug 4 - 5 16:03:23 Location: 9001 North I H 35 #150, Austin 16:03:29 #link https://www.eventbrite.com/e/magnum-midcycle-tickets-26245489967 Ticket 16:03:56 So far, I sold 11 tickets 16:04:14 Good sale! 16:04:17 those who plan to come, please RSVP so I can get a headcount for organizing the event. 16:04:35 Any way to attend for remote attendees? 16:04:54 yes, we'll try to use google hangouts like we did last time 16:05:32 Thanks muralia_ 16:05:37 muralia_: We have to do anything else than buying a ticket from eventbrite? 16:05:52 nope, just that. hongbin will give me an update 16:05:59 ok 16:06:36 OK. Advance topic 16:06:38 #topic Essential Blueprints Review 16:06:43 o/ 16:06:45 1. Support baremetal container clusters (strigazi) 16:06:51 #link https://blueprints.launchpad.net/magnum/+spec/magnum-baremetal-full-support 16:07:15 tcammann: Have you bought the ticket for midcycle? https://www.eventbrite.com/e/magnum-midcycle-tickets-26245489967 16:07:21 o/ I'm late! 16:07:41 I have updated the bp's whiteboard. Currently i'm working on setting up a gate job for a k8s ironic bay 16:08:21 Someone could start writing swarm-fedora-ironic templates 16:08:43 similar to k8s-fedora-ironic 16:09:25 I don't have anything else for this bp 16:09:42 strigazi: The k8s-fedora-ironic is working? 16:09:46 I would like to write the template for swarm 16:09:56 on devstack yes 16:10:03 thanks mkrai_ 16:10:22 is it possible to test on real baremetal yet? 16:10:26 hongbin, it's better to have a gate job to verify this 16:10:53 strigazi: The gate job is fine, but I guess poeple are looking for verifying it manually as well 16:10:55 I think it is, but you must have a real openstack deployment with ironic 16:11:29 strigazi: ack 16:11:43 Is there instruction on the ironic wiki? 16:11:49 (for real bare metal) 16:11:51 hongbin, having a job will help us verify that future patches won't break them and will force us updating them 16:12:14 tonanhngo, haven't check. I'll verify it 16:12:24 strigazi: I agree with that. My point is folks are earger to verify it manually :) 16:12:45 ok, I would be interested to try it on baremetal 16:13:02 tonanhngo ok 16:13:27 tonanhngo, let me what happen 16:13:30 the thing is 16:13:45 that you need at least two baremetal machines for that 16:14:04 right, because of our bay 16:14:15 I'll try to find as well 16:14:45 it's exciting to break bare metal and not only vms :) 16:15:34 OK. Any further comment? 16:15:48 2. Magnum User Guide for Cloud Operator (tango) 16:15:57 #link https://blueprints.launchpad.net/magnum/+spec/user-guide 16:16:22 The Swarm section merged, and the baymodel section is currently under review. Thanks everyone for the helpful reviews. 16:16:42 I am working on the bay section now, should have it uploaded in the next few days 16:16:54 so, steady progress. 16:17:00 that's all I have 16:17:11 Thanks tonanhngo 16:17:19 also, please take a look at the bay driver section patch everyone 16:17:36 That lead to the next BP 16:17:37 3. COE Bay Drivers (jamie_h, muralia) 16:18:23 muralia_: status? 16:18:34 so, this patch needs a little bit more work. since I submitted a new patch last week, a few new patches got merged to the old classes. I need to move those today. I'm seeing functional test failures currently and working on those. 16:19:31 do review it thought as its almost done. just needs some tweeking for the failures. 16:19:55 I've removed references to the old classes and moved over swarm related code and templates to the driver folder 16:19:59 thats it for me 16:20:13 muralia_: I have a question 16:20:17 yes 16:21:03 The driver is loaded dynamically? Does magnum searches for the available drivers? 16:21:46 no, not yet. first step it to move the code while still using the config file as the entry point. 16:21:57 ok 16:22:01 however, i've removed the hardcoded paths to the templates 16:22:21 those are discovered as we know the folder structure 16:22:23 I'll try to merge it to my patch and push for k8s 16:23:13 that's all 16:23:18 muralia_: A minor suggestion. Since you patch is large, you can do this 16:23:46 muralia_: When you patch is ready to merge, you can search the list of conflicted patches, and giva -1 on them to request the conflicting patches to merge later or depends on your patch 16:24:18 sure. i will do that. thanks 16:24:24 ok 16:24:44 Hopefully, that will reduce the amout of work to resolve conflicts 16:24:47 4. Create a magnum installation guide (strigazi) 16:24:54 #link https://blueprints.launchpad.net/magnum/+spec/magnum-installation-guide 16:24:58 Good news 16:25:00 http://docs.openstack.org/project-install-guide/container-infrastructure-management/draft/ 16:25:35 I also pushed and tested with the latest opensuse packages 16:25:40 https://review.openstack.org/#/c/335051/ 16:26:34 WIP: build deb packages to test on ubuntu/debian from master and lbaasV1 install notes 16:26:47 that's it 16:27:13 strigazi: Thanks. This one looks professional :) http://docs.openstack.org/project-install-guide/container-infrastructure-management/draft/ 16:27:32 strigazi: Thanks for the hard work 16:27:40 :) 16:27:51 strigazi: Thanks. Looks really cool :) 16:28:25 #topic Magnum UI Subteam Update (bradjones) 16:28:37 I didn't receive any update from brad 16:28:47 So let's skip this 16:28:59 #topic Kuryr Integration Update (tango) 16:29:19 tonanhngo: ^^ 16:29:27 I attended the Kuryr meeting yesterday 16:30:01 most of the discussion was on the code refactoring, which will impact our integration 16:30:40 Then, we need to wait for them to finish? 16:30:42 basically, the current code is specific to libnetwork, so it needs to be generalized to accommodate CNI and other drivers 16:31:13 no, I am moving ahead with the current code, but our patches will remain WIP until they finish the refactoring 16:31:28 ack 16:31:33 So we will make adjustment along the way as they proceed 16:32:07 I am still debugging the Swarm bay with Kuryr 16:32:53 I tried the container image for openvswitch agent from Kolla. It needs some config so I am working on this 16:33:15 I may need to ping the Kolla team if I get stuck 16:33:57 I also asked Vikas to update the Kuryr spec on the part mentioning Magnum API for container 16:34:08 He will remove those references 16:34:30 That's all for now 16:34:37 Thanks tonanhngo 16:34:47 Any question for Ton? 16:35:13 #topic Other blueprints/Bugs/Reviews/Ideas 16:35:19 1. Asynchronous Operations for bay? 16:35:24 #link https://blueprints.launchpad.net/magnum/+spec/async-container-operations Asynchronous Container Operations 16:35:35 muralia_: I remembered you requested to discuss this one? 16:35:53 yes, there are a few patches out there for async bay apis 16:36:16 it changes the create behavior a little but 16:36:44 the api does not return the entire bay object now, just the uuid as its still being created when the api returns 16:37:06 what does this affect? 16:37:14 will it affect the UI integration? 16:37:24 just wanted everyone to be aware 16:37:38 I am not sure the UI part, need to talk to Shu about that 16:37:56 ok. 16:38:08 hongbin is Shu able to attend our team meetings, or is that not practical because of timezones? 16:38:30 adrian_otto: He is in Japan, it is late everning for him 16:38:40 Link to Async Bay Operations patch: https://review.openstack.org/#/c/330771/ 16:38:41 I see 16:38:58 we'll need to update the getting started docs. Can you folks think of anything else this might impact? do leave comments 16:39:49 muralia_: which docs do you mean? 16:39:51 muralia_: Heat has a bay resource. Might worth to check if it is still working after teh patch 16:40:18 muralia_: mkrai_ Murano has a Magnum plugin, does it still work? 16:40:36 strigazi: actually, i just checked again and the getting started guide does not show what bay create returns. so we are good. 16:40:47 hongbin: I see some error on Magnum App 16:41:01 But haven't looked in detail 16:41:10 hongbin: thanks. we'll check all of these 16:41:26 mkrai_: ack. Does it related to the aysnc patch or not? 16:41:36 I guess it is not 16:41:46 It is some authorization error 16:41:52 mkrai_: OK. We can work on that offline 16:42:01 Sure 16:42:29 Any other comment on this topic? 16:42:34 hm, somebody says 'Murano' here 16:42:55 hi folks :) 16:43:11 freerunner: Hi, we talked about the magnum-app for murano 16:43:13 freerunner: Yes we are talking about Magnum app 16:43:16 Looks like we have a bug for your case, yep. https://bugs.launchpad.net/murano-apps/+bug/1595511 16:43:16 Launchpad bug 1595511 in murano-apps "[Magnum][Murano] deployment of Magnum bay application is failed" [High,Confirmed] - Assigned to Madhuri Kumari (madhuri-rai07) 16:43:51 freerunner: Thanks for it. I will try to submit the fix tomorrow 16:44:19 mkrai_: Wow! Thank you very much! 16:44:25 freerunner: thanks for that. we were also discussing a new change to make bay API's async. 16:44:41 it changes the bay create behavior a little bit. 16:45:12 it now only returns a uuid as the bay is being created in the background. 16:45:37 https://review.openstack.org/#/c/330771/ 16:45:59 muralia_: Thanks for putting this up. I will check it for Magnum app 16:46:21 freerunner: can you take a look at this patch and comment on it. we'd like to know if it breaks anything for you 16:46:54 #topic Open Discussion 16:46:57 muralia_: Will try to review it soon :) 16:47:02 thanks 16:47:48 Anyone has a topic to discuss? 16:47:57 yeap 16:48:05 strigazi: go ahead 16:48:36 Do we have a list of magnum users? Do we now people using it? 16:49:10 Can we do something about it? 16:49:41 Is it possible to have section in the wiki page for that? 16:49:49 strigazi: good idea 16:49:52 that's it 16:50:03 strigazi: I believe CERN is a user of Magnum? 16:50:27 yes we have two blog posts about our use case 16:50:50 Yes, from what I know, CERN is the only reported user 16:51:28 muralia_: adrian_otto Does Rackspace have any plan to adopt Magnum? 16:52:19 yes, we have internal deployments, but not available to our users yet. we do plan to. 16:52:38 muralia_: ack 16:52:50 I will create a wiki session for listing the users 16:53:06 For everyone, you can add yourselves if you want 16:53:24 I guess we will have solicit and get agreement for the users to be listed 16:53:24 #action hongbin create a wiki session to report a list of Magnum users 16:53:48 tonanhngo: Yes, it is up to them to edit the wiki to list themselves 16:55:13 OK. If there is no other topic, we can finish a bit earlier 16:55:30 #endmeeting