17:00:15 #startmeeting ironic 17:00:16 Meeting started Mon Jul 3 17:00:15 2017 UTC and is due to finish in 60 minutes. The chair is TheJulia. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:17 o/ 17:00:18 o/ 17:00:18 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:20 The meeting name has been set to 'ironic' 17:00:30 o/ 17:00:36 o/ 17:00:37 \o 17:00:41 o/ 17:01:17 o/ 17:01:28 This being a holiday week, I'm expecting attendance to be light today. 17:01:40 \o 17:01:44 Our Agenda can be found on the wiki 17:02:00 o/ 17:02:04 o/ 17:02:05 o/ 17:02:06 #link https://wiki.openstack.org/wiki/Meetings/Ironic 17:02:49 #topic Announcements / Reminder 17:02:57 o/ 17:03:00 o/ 17:03:46 Morning!!! 17:03:57 o/ 17:04:18 First off, Our documentation on docs.o.o is broken. sambetts has started putting together changes to hopefully address this for our main repo, but all of our doc tress will likely require some attention over the next few days. I'll pickup on sambett's work after he decides to call it a night. 17:04:40 thanks TheJulia :) 17:04:43 This is a result of the project wide documentation migration project. 17:04:50 o/ 17:04:57 #link https://etherpad.openstack.org/p/doc-migration-tracking 17:05:29 sambetts: I'll review them against the spec/docs/instructions as well, hopefully we can get the majority of this cleaned up quickly 17:05:46 Secondly, milan wanted to let us know that we have a new person joining us. 17:05:54 TheJulia, thanks! :) 17:06:10 Greetings etingof! Welcome to this crazy batch of people :) 17:06:20 hey etingof 17:06:21 I'm very happy to introduce Ilya Etingof, etingof, to the Ironic folks! 17:06:38 thank you! I'm really excited joining you! ;-) 17:06:59 welcome :) 17:07:41 hey etingof, welcome 17:07:48 etingof: Please don't hesitate to ask questions. We've all learned this is a very complex project, so don't feel bad if you have questions, or need help at any time. 17:08:02 TheJulia, ++ 17:08:15 sure, thank you! 17:08:17 With that having been said, does anyone else have any announcements/reminders for us? 17:09:15 * TheJulia guesses no 17:09:57 Moving on! 17:10:04 #topic Subteam status report 17:10:12 #link https://etherpad.openstack.org/p/IronicWhiteBoard 17:10:29 Starting at line 92... 17:13:13 * TheJulia sense people typing furiously :) 17:15:23 so... 17:15:57 It looks like we didn't get as much as we hoped done last week, and it also looks like some of the sections were not updated since people with context are jut not present 17:16:31 * TheJulia is running a little slow because of connectivity 17:17:49 I don't have any questions, and I guess we can just move forward to priorities for the week if nobody has questions/comments from the ether pad? 17:19:46 milan: If that is you typing, yes, we've all been impacted by the docs changes :( 17:20:05 TheJulia, ack, yeah wanted to record it 17:20:11 * milan files a bug 17:20:18 * for inspector 17:20:41 or do we have some template for that? 17:20:56 almost forgot 17:21:01 o/ 17:21:19 milan: the changes, see the ether pad for the doc migration. It has a link to a spec with instructions. :) 17:21:39 \o NobodyCam 17:21:40 TheJulia, ack, thanks :) 17:21:47 milan: no problem. :) 17:22:06 I guess we should move on 17:22:11 #topic Deciding on priorities for the coming week 17:22:37 I think the first one has to be docs right? 17:22:56 Yup! 17:23:06 I think that is one we're going to have to update as we go 17:24:04 \o/ kicked off etherpad 17:24:27 and it is still loading 17:24:54 Second, I think would be BFV. We have two patches before we can begin unblocking the client API patches. Both of the are in the list on the etherpad 17:25:49 It looks like the rest is fairly reasonable since it will be a relatively quiet week. 17:26:08 Does anyone have anything specific they wish to raise to that list? 17:26:50 TheJulia: can we test BFV already? :) 17:27:25 xavierr: Yes, There is a notes ether pad covering that someplace. I'll have to look 17:27:29 it would be nice to get the physnet awareness work back on the priority list. 17:27:34 +1 17:28:09 there are a few patches backing up, some only tangentially related to physnets but seem to have ended up blocking it 17:28:30 First blocker is https://review.openstack.org/477597 17:28:45 mgoddard_: sambetts|afk Is there anything that is needed in nova for it? 17:28:53 TheJulia: ty, paste the link if you find it :) 17:29:00 TheJulia: no nova changes required 17:29:30 mgoddard_: nothing neutron/client library related that needs to land this cycle? 17:30:02 TheJulia: nope, it's all in ironic-governed projects 17:30:04 mgoddard_: do we need to add physical network to the client?? 17:30:29 The field I mean, to like port create etc 17:30:32 and does it need to occur this cycle for the client library release freeze? 17:30:33 sambetts_: https://review.openstack.org/#/c/461893 adds physnet support to ironic OSC commands 17:30:52 when's the client release freeze? 17:31:28 double checking 17:31:44 mgoddard_: Jul 24 - Jul 28 17:31:50 https://releases.openstack.org/pike/schedule.html 17:32:08 no pressure then :) 17:32:20 I was just going to paste that :) 17:32:22 #link https://releases.openstack.org/pike/schedule.html 17:32:44 * xavierr was faster 17:32:56 (or you just have less lossy internet) 17:32:57 :) 17:33:10 hahaha 17:34:36 3 weeks. sambetts_, thoughts on parking the VIFPortIDMixin refactor in favour of pushing forward physnet support before feature freeze, which is apparently 4 weeks away? 17:35:03 We really need to get the stuffs in for nova for BFV this cycle, I'm personally okay if we cut >1 client library releases leading up to that date. Personally I want to get the bfv stuff in a place where we have unblocked nova and can cut a client library release, and without dmitry, I'm hesitant to re-add it to the priorities list. That being said, it sounds like the blocker needs to be resolved first? 17:35:43 The key is client library freeze, we'll just become hesitant to land code to ironic the closer we get to when we need to cut a release, as we always do. 17:36:09 which patches do we need to unblock nova? only the rest api one? 17:36:22 it's really a soft blocker, in that an early code refactor makes later patches cleaner 17:36:29 oh, and client 17:37:04 vdrok: two for the rest api to land, two client patches to which allows targets to be added if memory serves. 17:38:57 mgoddard_: can we make the field accessible on the API before we start using it in vif port mixin? That unties us from the client libs 17:40:11 sambetts_: it's possible, just means we end up with a behaviour change without a corresponding API bump 17:40:39 That sounds a little dicey 17:40:43 Or we land messy vif port ID mixin with the promise that we clean it up before any more code gets landed in vif port ID mixin 17:40:53 I guess we could later bump again if that is an issue 17:41:46 Making vif attach port attach pay attention to physical network doesn't require a bump 17:42:22 it shouldn't as long as behavior can stay consistent for the users 17:42:26 In the same we we didn't bump when we enable post deploy attachmeny 17:42:56 ok, if there's a precedent then I guess it's ok 17:42:58 But this is a grey area for API bumping 17:43:46 Because it's actual driver functionality leaking through to the API 17:43:58 Yeah, given the timeframe, I'm okay with minimum viable path to get the client api patch landed and then doing the refactoring. Dmitry may feel otherwise though, so as we say in the states, your milage may vary. 17:44:08 I'd prefer both field and its use to be landed together, but yeah, it is possible to do separately 17:45:05 Ok then I say land messy patch with all the extra checks and follow it immediately with the vif port mixin refactoring 17:45:06 It seems like we need to get a plan together for that, and we should be able to move forward on that if we can also get bfv moving forward this week 17:46:00 sambetts_: +1 17:46:41 * mordred waves 17:47:00 Sounds like we are in an okay shape for this week, and we can move forward to open discussion, that is if there are no objections? 17:47:09 Sounds good 17:47:19 morning mordred :) 17:47:26 #topic Open Discussion 17:48:05 Anyone have anything else to discuss today? 17:48:27 Hi, regarding this patch: 17:48:27 https://review.openstack.org/#/c/358142/ 17:48:28 TheJulia, hi. 17:48:49 This has been reviewed by osprofiler cores several times. Needs Ironic cores' review. 17:49:49 rama_y: We have very few cores on hand today, but we can make a note of it? 17:49:53 regarding https://review.openstack.org/#/c/408441/ 17:50:02 TheJulia, thanks very much. 17:50:43 #note rama_y is seeking core reviewers to review the osprofiler addition https://review.openstack.org/#/c/358142/ 17:51:11 * mordred can't remember who was working on the "consume endpoints from catalog" patches -but I wanted to let folks know that the keystoneauth patches to make that all work really well are mostly landed 17:52:01 we've landed list-of-interfaces, support for major version ranges and reporting discovered microversion support - hopefully will land service-type alias support this week or next week 17:52:07 pas-ha was, I'll let him know, thx! 17:53:14 This patch as i understand is pending in lack of CI. We at HPE is lacking hardware to get this feature onto CI. I sit OK if we have it in tempest as of now and support it in CI when we get the hardware. 17:53:20 https://review.openstack.org/#/c/408441/ 17:53:45 s/I sit/Is it 17:54:06 hrm , isn't it just a doc patch? 17:54:12 * Nisha_Agarwal just asking opinion 17:54:25 yes 17:54:45 oh, I see the comment from JayF 17:55:32 \o/ irccloud 17:55:39 Anyway! 17:56:11 lol 17:56:18 Nisha_Agarwal: I think reviews on doc changes will need to wait unless it is on a critical path for client library release 17:56:58 client library includes third party libraries? 17:57:22 mordred: I'll pass that along to pas-ha 17:57:39 * Nisha_Agarwal is always confused with these terminologies 17:57:56 Ironic's core reviewer bandwidth needs to focus on making sure python-ironicclient can be released in time for it to be consumed by other projects 17:58:05 ok 17:58:12 With that being said, we have two minutes left 17:58:17 Anything else today 17:58:18 i understand that part and agree also 17:58:31 nothing else from my side 17:58:35 :) 17:58:40 TheJulia, regarding the osprofiler patch, rloo has provided significant feedback on the patch and the patch has been revised; just wanted to add to my earlier request. 17:58:45 Thanks for clarifying 17:58:57 regarding the client, would be also good to have some reviews on https://review.openstack.org/359061 18:00:13 Yeah, we do need to review that 18:00:15 thank you vdrok 18:00:17 * TheJulia makes a note 18:00:22 thanks all! 18:00:32 #note vdrok points out that we need reviews on the python-ironicclient patch https://review.openstack.org/359061 18:00:43 Thanks everyone! See you next week, and in #openstack-ironic in the mean time 18:00:47 #endmeeting