16:00:20 #startmeeting containers 16:00:21 Meeting started Tue Aug 2 16:00:20 2016 UTC and is due to finish in 60 minutes. The chair is hongbin. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:22 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:24 The meeting name has been set to 'containers' 16:00:25 #topic Roll Call 16:00:32 o/ 16:00:37 Ton Ngo 16:00:44 o/ 16:00:50 Spyros Trigazis 16:00:59 o/ 16:01:06 o/ 16:01:55 Thanks for joining the meeting muralia_ tonanhngo Drago strigazi dane_leblanc_ rpothier 16:02:06 o/ 16:02:09 Adrian Otto 16:02:10 #link https://wiki.openstack.org/wiki/Meetings/Containers#Agenda_for_2016-08-02_1600_UTC Today's agenda 16:02:18 Anything needs to be added to the agenda? 16:02:43 #topic Announcements 16:02:50 1. Welcome Spyros Trigazis to join the Magnum core team 16:02:58 #link http://lists.openstack.org/pipermail/openstack-dev/2016-July/100455.html 16:02:59 * adrian_otto applause 16:03:00 woohoo! congrats 16:03:13 Congrats strigazi! 16:03:15 well deserved 16:03:25 Thanks you :) 16:03:29 Congrats strigazi 16:03:43 Congrats strigazi 16:03:44 Thanks strigazi for your significant contribution 16:03:48 Congrats Spyros! 16:04:03 2. Vote for mascot ended. The team mascot is sharp. 16:04:09 #link http://civs.cs.cornell.edu/cgi-bin/results.pl?id=E_15e176081cd9d2b1 16:04:28 I sent the choice to the OpenStack foundation 16:04:41 If no other team picked this mascot, we will get it 16:04:45 hongbin: did we also send waves as an alternate? 16:04:50 oh, ok 16:04:52 yes 16:05:16 #topic Review Action Items 16:05:18 None 16:05:26 #topic Essential Blueprints Review 16:05:31 1. Support baremetal container clusters (strigazi) 16:05:37 #link https://blueprints.launchpad.net/magnum/+spec/magnum-baremetal-full-support 16:05:42 strigazi: ^^ 16:06:17 Not much this week, we are looking into creating a CI for testing magnum and ironic on our infra 16:06:34 we might open it for everyone if it goes well 16:07:01 with devstack vms with more memory 16:07:43 strigazi: How about the CI with OpenStack infra? 16:07:45 That's it, last week we did some benchmarks for magnum with rally. I'll tell you on the midcycle 16:08:15 hongbin: what do you mean? 16:08:46 strigazi: Maybe to clarify, are you setting up a third-party CI or a pipeline in OpenStack Infra 16:09:33 third-party CI like in other projects, but it's not final 16:10:08 in openstack infra it seems that we have memory limitations 16:10:40 strigazi: We cannot work around the memory limitation? 16:11:19 we'll try. In one of the tests the vm was swapping 16:11:39 I see 16:12:14 Thanks strigazi for working on this 16:12:21 Any question for strigazi ? 16:12:55 2. Magnum User Guide for Cloud Operator (tango) 16:13:00 #link https://blueprints.launchpad.net/magnum/+spec/user-guide 16:13:05 tonanhngo: ^^ 16:13:07 The TLS section was merged. 16:13:30 The bay section is still under review 16:14:41 I don't have new section today 16:15:59 tonanhngo: ack. anything else? 16:16:12 Thanks to Spyros and Ricardo for great feedback 16:16:31 https://review.openstack.org/#/c/340670/3/doc/source/userguide.rst@327 16:17:48 tonanhngo: Do you think that is unecessary? For example last week we were creating bays in parralel, with different requests of course 16:18:08 Not sure what you mean 16:19:19 A user doesn't have to wait for a bay to have status CREATE_COMPLETE 16:19:44 ah ok, I will update that part 16:20:03 Thanks for the observation 16:20:19 thanks 16:20:20 Thanks tonanhngo 16:20:31 3. COE Bay Drivers (jamie_h, muralia) 16:20:36 #link https://blueprints.launchpad.net/magnum/+spec/bay-drivers 16:20:39 muralia_: ^^ 16:21:07 I don't have any updates since last week. i've been on vacation for a few days and just got back. I will continue working on my patch 16:21:37 I have a question for everyone 16:21:50 Are we going to remove the replication controller object? 16:22:02 yes we should. 16:22:11 ok, I'm almost tehre 16:22:15 ok, I'm almost there 16:22:16 thanks 16:22:40 Thanks muralia_ . Any other question? 16:23:02 #topic Kuryr Integration Update (tango) 16:23:12 tonanhngo: ^^ 16:24:16 We hit a little snag last week. The Kuryr image is broken because it got rebuilt using the refactored repos, which are not complete 16:25:44 I told the Kuryr team, they will build and tag a stable image 16:25:46 tonanhngo: that's one of our images, or something made by the Kuryr team? 16:25:55 It's by Kuryr 16:25:56 I see 16:26:57 Thanks tonanhngo . Anything else? 16:27:16 That's all for now 16:27:20 Thanks 16:27:29 #topic Other blueprints/Bugs/Reviews/Ideas 16:27:35 1. Midcycle topics 16:27:41 #link https://etherpad.openstack.org/p/magnum-newton-midcycle-topics 16:28:02 I put all the topics to the agenda 16:28:28 There are three slots left, which we can use to discuss alternative topics 16:28:47 I invited strigazi to give a presentation at the opening 16:29:08 strigazi: will share the experience of using Magnum at his group 16:29:45 One more topic: Add local (configurable) docker registry 16:29:57 At the end of the first day, we could plan a team event (drinks or dinner or others) 16:30:10 strigazi: ack. feel free to add it to the etherpad 16:30:15 added 16:30:40 Great to see the agenda filled out. Excited to have everyone here. 16:30:56 muralia_: adrian_otto Do you have any arrangement for a team event (drinks? dinner?) 16:30:59 enjoy, pity i can't make it but next time :) 16:31:17 yes, we were thinking about drinks at the rackspace office friday evening 16:31:31 hongbin: lets get that discussion opened with the midcycle attendees 16:31:39 ok 16:31:48 thursday evening there is a community meetup 16:32:01 all team members are welcome, and encouraged to attend 16:32:16 link to the meetup? 16:32:19 http://www.meetup.com/Docker-Austin/events/232067917/ 16:32:36 Awesome 16:32:37 weird how out mid-cycles fall on docker meetup days 16:32:43 our* 16:32:48 :) 16:32:48 #link http://www.meetup.com/Docker-Austin/events/232067917/ 16:33:42 Any other question / comment for the midcycle? 16:34:17 2. Microversioning implementation 16:34:23 #link https://review.openstack.org/#/c/343060/ 16:35:03 jvgrant: ^ 16:35:07 The patch above is a proposal to implement a decorator for specifying microversion 16:35:37 Eli Qiao gave a -2 with comments that he don't like the decrorator approach 16:35:53 this provides a common framework and style versioning, 16:36:32 currently it would be checking and making changes within each api method 16:36:59 this really is just providing a more structured versioning style 16:37:29 it based on Nova's way of doing it using a decorator 16:37:43 yes, i prefer the decorator approach for versioning. we'll be using it a lot more when we implement lifecycle operations etc. 16:37:59 I also prefer the decorators. 16:38:05 +1 decorators 16:38:05 me as well 16:38:13 +1 decorators approach 16:38:26 Any opposing point of view? 16:38:30 is Eli in attendance today? 16:38:38 eliqiao: ^^ 16:38:45 +1 decorators approach 16:38:54 I guess he is offline since it is late everning for him 16:38:55 i think Eli's main concern was the additional complexity 16:39:19 well he used a -2 vote, which we need to consider carefully 16:39:41 I'm reluctant to support an alternate decision without a conversation about it 16:39:50 even though I don't agree with him 16:40:09 I will ping him later for discussing it further 16:40:13 thanks hongbin 16:40:21 thanks hongbin 16:40:34 #action hongbin discuss with eliqiao for the microversion implementation 16:40:35 Does anyone know if he will be at the midcycle? 16:40:43 No, he won't 16:40:58 He didn't sign the event. 16:41:23 and flighting from China to US is $$$ 16:41:43 can we arrange a conference call at a convenient time for him? 16:41:43 I guess it is very hard for him to get approval 16:42:09 I think we all can relate. Okay, let's do our best to consider his POV. 16:42:51 OK. Then, I can talk to him to schedule a call 16:43:36 Anything else for this topic? 16:44:26 #topic Open Discussion 16:45:24 If nothing else to discuss, let's end the meeting a bit earlier 16:45:38 i have a question on the param names for the client 16:45:41 with all the -id 16:45:59 there's a patch to drop it in external-network-id. could we do it for all? it's confusing 16:46:09 maybe adding new ones and flagging the old ones as deprecated 16:46:20 or else we do it for the osc only 16:46:42 I think flagging the old ones as deprecated is the correct approach 16:46:56 Just renaming will break backward-compatibility 16:47:06 sure 16:47:55 rochaporto: is your concern that fact that some attributes contain the string "-id" and others do not? 16:48:36 yes, and the -id itself being there 16:48:43 given that it can take the name too 16:49:09 Definitely the *-id suffix should be removed 16:49:09 early on it only accepted an id 16:49:22 @rochaporto are your taliking about: https://bugs.launchpad.net/python-magnumclient/+bug/1599393 16:49:23 Launchpad bug 1599393 in python-magnumclient "magnum baymodel-create optional parameter(external_network_id) needs update" [Undecided,In progress] - Assigned to yatin (yatinkarel) 16:49:25 we added the code to accept names as a convenience a bit later 16:49:38 This should improve the user experience 16:49:53 ok, the issue is clear to me now, thanks. 16:49:59 yatin: yes, but same is true for other params 16:50:22 small thing, but might as well get it done 16:50:22 Yes, i commented on the bug itself 16:50:50 yatin: Are you the preson who working on fixing the bug? 16:50:58 Yes 16:51:18 yatin: Do you have enough instruction on how to proceed ? 16:52:22 yatin: For example, how to flag an attribute as deprecated, and add the new one .... 16:52:53 hongbin: no, but i will try to find out. 16:53:04 if you approve this 16:53:21 yatin: ok. ping the #openstack-containers if you need help 16:53:37 yatin: Thanks for volunteering to fix this bug 16:53:50 Ok 16:54:28 thanks :) i'll look into other params in parallel 16:55:13 All, thanks for joining the meeting today. The Magnum midcyle will hold at this Thursday & Friday. Hope to see you all there 16:55:20 #endmeeting