02:37:46 #startmeeting openstack-cyborg 02:37:46 Meeting started Thu Jul 15 02:37:46 2021 UTC and is due to finish in 60 minutes. The chair is xinranwang. Information about MeetBot at http://wiki.debian.org/MeetBot. 02:37:46 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 02:37:46 The meeting name has been set to 'openstack_cyborg' 02:37:55 #topic roll call 02:38:13 #info brinzhang 02:38:17 #info xinranwang 02:38:47 #info songwenping__ 02:39:00 here is our agenda, the first topic is PTG plan. https://wiki.openstack.org/wiki/Meetings/CyborgTeamMeeting#Agenda 02:39:09 #topic PTG plan 02:39:59 the ptg registration is open now. do you thik 6-8 UTC time(14-16 Beijing Time) is good for you 02:40:05 today will free specs 02:40:23 brinzhang: yes, that's the next topic 02:40:36 ah, follow you, is there some specs need to be done today? 02:41:02 ptg time is ok for me 02:41:02 let's talk about this later. 02:41:21 I am not sure that time whether is ok for me, it too far from now 02:41:23 we need have an agreement about the ptg time slot 02:41:42 each time should register the slot now 02:42:05 hi all 02:42:07 when ptg begin? 02:42:12 yeah, I agree. it is far from now... 02:42:28 Oct 18th I remember 02:42:28 hi arthurdayne 02:42:31 welcome to join 02:43:16 it's really far 02:43:28 brinzhang: do you think the evening time is better? 02:43:31 I think we can decide it as your suggestion, if there need to be changed we can try to change 02:43:50 songwenping_: yeah, I don't know why it starts such early this time. 02:44:03 except the topic need to be cross many project team ^^ 02:44:19 brinzhang: ok, so let's make it as 6-8 UTC 02:44:39 and we plan to have a cross project with nova 02:44:46 to discuss the vgpu support 02:45:12 which need your presence as well 02:45:20 about the cross-project topic we shuold follow nova team's time, so I think it will be later for us 02:46:02 I will send a maillist to nail down the time slot about cross project 02:46:30 about the duration, do you think 1 hour is enough? 02:47:00 the vgpu's controversial point is the trait name, may 1 hour is enough, do you think so? 02:47:33 yeah, do you have any other topic want discuss with nova? 02:48:14 no, and I have another nova topic, there is no time to update, so it is pending long time :( 02:48:59 never mind, so we only have vgpu topic right? 02:49:11 from now, I think yes 02:49:21 ok, great 02:49:56 #topic spec freeze day 02:50:12 next topic is about spec freeze. 02:50:29 the spec will be freezed at the end of today . 02:50:43 for now, we have 3 pending specs. 02:51:06 first one is PGPU spec 02:51:44 sorry for missing the previous spec submission, I think we should address the previous comment as well. 02:51:53 what do you think songwenping_ ? 02:53:01 but yumeng can't review again 02:53:10 it's all yument's comments 02:54:56 I agree with yumeng about adding more device specific info in the spec, like data model, rp structures etc, it is good. 02:56:27 ok, i'll add these infos today. 02:56:59 I mean we should use the old spec, there are some discussion we can trace 02:57:10 cool. thanks songwenping_ for the effort. 02:57:37 brinzhang: both are ok for me. 02:57:47 so I would like you can restore the old one, and update or copy your updated info in it 02:58:26 maybe no need to restore, make the old one as a reference is ok for me 02:59:14 and we can add the new spec link as the abandon message in previous one. 02:59:28 yes, but it looks not a good operation 02:59:55 the goal is to make it well traced, you can choose the way you prefer. 03:00:21 if you are all agree, I can follow ^ 03:01:39 brinzhang: thanks brinzhang 03:02:28 songwenping_: seems you have dropped off, we were discussing your PGPU spec, please check the log and let us know if any question. 03:02:46 the next pending spec is pmem spec. 03:03:05 no question, please go on 03:03:09 is fossen around? 03:03:19 he is offline 03:03:24 i ping him 03:03:36 he is off seat 03:03:42 ok, we well discuss it when he's back. 03:03:53 ok let's go on 03:03:57 the third one is disable/enable api spec. 03:04:20 I think this one will be postponed to next release. 03:05:04 it lacks of reviews and updates. I don't think we can land it by today... 03:05:25 what do you think 03:06:56 where is the spec? 03:07:33 https://review.opendev.org/c/openstack/cyborg-specs/+/790917 03:07:42 https://review.opendev.org/c/openstack/cyborg-specs/+/799173 03:07:58 if you have time to review, I can update it today and try to land it :) 03:08:03 hi qiujunting 03:08:10 thanks for your spec 03:08:43 i am sorry i miss this spec 03:09:11 xinranwang:hi 03:09:33 could you please add more details in the spec, I have some comments on it. 03:09:45 I am sorry, i will update the spec later 03:09:45 there are too many incoming reviews in my list. 03:10:24 qiujunting: it's ok. so we may need to postpone the pmem spec to next release 03:10:48 xinranwang: will review today, i think this is not complex. 03:11:47 ok 03:12:32 qiujunting: from my perspective, we need some introduction about the pmem, which mode we use, why we don't support other mode, etc. which kind of resouece vm treat it as, etc 03:12:52 songwenping__: ok, thansk, big core review ;) 03:13:20 ^-^ 03:13:38 ok, i will update this to the spec 03:14:11 so that's all status of the pending spec. 1 spec postponed, and 2 specs trying to be merged by today. 03:14:38 qiujunting: ok, thanks a lot 03:15:24 #topic vgpu support 03:15:37 the next topic is vgpu support. 03:16:01 our env is not available 03:16:14 why 03:16:26 still the license problem? 03:16:32 no 03:17:50 my server cann't visit extranet 03:18:21 our company is perfoming network check 03:19:05 is there any tunnel permit you get access to extranet 03:19:08 maybe next week the application will be passed 03:19:30 this need apply 03:19:41 ok, so it is a temporary problem 03:19:53 i think the code is ok mainly 03:20:01 i have tested before. 03:20:56 yes, I remember. 03:21:06 https://etherpad.opendev.org/p/nova-yoga-ptg 03:21:29 we need the detailed test report. 03:21:30 here is nova's etherpad, please add the vgpu discussion on it. 03:21:51 ok thanks 03:23:10 songwenping__: yeah. But nova has some different opinion, I think we should have an agreement first. we can show them the test report of current implementation, it is nice to have. 03:23:26 sure 03:24:50 ok,cool 03:25:23 #topic AoB 03:25:34 so please review this patch: https://review.opendev.org/c/openstack/cyborg/+/790937 03:25:47 and there is nothing from myside 03:26:20 yeah, I was planning to discuss it during meeting, but we are running out of time. I will leave the comments 03:26:48 ok, if nothing else, let's wrap up the meeing. 03:26:56 yeah 03:27:15 please remember we have 2 spec need review and updates by today! 03:27:21 bon courage 03:28:07 #endmeeting