14:00:17 <zhaochao> #startmeeting trove 14:00:18 <openstack> Meeting started Wed Apr 18 14:00:17 2018 UTC and is due to finish in 60 minutes. The chair is zhaochao. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:19 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:21 <openstack> The meeting name has been set to 'trove' 14:00:31 <zhaochao> #topic roll call 14:01:05 <gang> hi 14:01:18 <zhaochao> gang: hi 14:01:53 <zhaochao> looks like today we only very few people to the meeting... 14:02:06 <gang> yes... 14:02:16 <kumarmn> hi 14:02:35 <zhaochao> wait some minutes for others to show up 14:02:38 <zhaochao> kumarmn: hi 14:03:25 <fanzhang_> hi sorry for a little bit late 14:04:01 <zhaochao> fanzhang_: :) 14:05:41 <zhaochao> OK, let's start 14:05:57 <zhaochao> #topic Rocky goal updates 14:06:58 <zhaochao> This week we're on rocky milestone-1, so I may have to submit some patches for it 14:07:50 <zhaochao> in the first and a half months of rocky cycle, we may don't have many changes 14:08:12 <fanzhang_> There are some patches from you, and great work! 14:08:55 <zhaochao> fanzhang_: thanks 14:09:28 <fanzhang_> Should we approve the nova file injection bp before rocky milestone-1? Maybe I can start developing ? 14:09:30 <kumarmn> zhaochao: You do have to submit a patch to the openstack/releases repository 14:09:32 <kumarmn> https://github.com/openstack/releases 14:09:56 <kumarmn> Take a look at the README. 14:10:20 <zhaochao> kumarmn: thanks, I'm just going to ask, anything more about the milestone ? 14:10:30 <zhaochao> fanzhang_: I think you could start developing 14:10:52 <kumarmn> something like 'tox -e venv -- new-release rocky trove milestone' 14:11:11 <fanzhang_> zhaochao sure, I guess I have to code step by step :) 14:11:24 <kumarmn> That will take care of trove and trove-dashboard. I believe for python-troveclient, you have to create the snapshot yourself. 14:11:29 <zhaochao> fanzhang_: I approved the bp, and will wait for another week for the other core members to approve it 14:12:17 <zhaochao> kumarmn: where should I run the tox command , in trove repo or the releases repo ? 14:12:33 <kumarmn> releases repo 14:12:41 <fanzhang_> zhaochao thanks :) really appreciate that anyone can help review this spec :https://review.openstack.org/#/c/557583/ 14:12:48 <zhaochao> kumarmn: got that, thanks 14:13:18 <zhaochao> #action zhaochao have to do the tasks about Rocky milestone-1 14:13:44 <gang> fangzhang: so, need a workflow+1? 14:13:58 <fanzhang_> #action fanzhang_ should start coding about nova file injection spec. 14:14:25 <fanzhang_> I'd like you actually review this spec first :) 14:14:28 <zhaochao> and for the global goal, mox-removal 14:14:42 <zhaochao> it's already solved in the Trove repo 14:14:52 <gang> of course :) 14:15:00 <zhaochao> and the other repo have to be dealed is trove-dashboard 14:15:11 <zhaochao> we now have https://review.openstack.org/#/q/project:openstack/trove-dashboard+topic:mox-removal 14:15:41 <fanzhang_> ok, will do review later 14:15:54 <zhaochao> 5 patches for this topic in trove-dashboard 14:16:30 <zhaochao> one of them from Chuck Short, but no updates recently to address my comments, so I also updated it today 14:16:34 <zhaochao> fanzhang_: thanks, :) 14:17:14 <zhaochao> and https://review.openstack.org/#/c/562144/ 14:17:35 <zhaochao> this one is related, without it, we cannot merge any patches... :( 14:17:52 <fanzhang_> better to add them to # link 14:18:16 <zhaochao> fanzhang_: good point, I'll do that 14:18:26 <zhaochao> #link https://review.openstack.org/#/q/project:openstack/trove-dashboard+topic:mox-removal 14:18:40 <zhaochao> #link https://review.openstack.org/#/c/562144/ 14:19:02 <zhaochao> another patch need to mention from me 14:19:06 <zhaochao> https://review.openstack.org/#/c/562144/ 14:19:10 <zhaochao> #link https://review.openstack.org/#/c/562144/ 14:19:38 <zhaochao> I'm going to migrate the zuul jobs to Zuul v3 native 14:20:21 <zhaochao> the above is the first one, the jobs of trove repo will go after 14:21:53 <zhaochao> I have 10(or 11) patches recently, please help to review, thanks 14:22:08 <gang> zhaohcao: must say you really have done a lot of work! 14:22:14 <zhaochao> that's all from me 14:22:28 <zhaochao> gang: :) 14:23:23 <zhaochao> one more thing about the stable branches, you guys may have seen the mails in the mailinglist about this 14:23:37 <fanzhang_> Yes, thanks to Sean and Matt 14:23:49 <fanzhang_> we got the patches merged 14:24:30 <zhaochao> Yes, I think as we don't have much resouce, so currently we will rely on the core stable team to help us to got the patches merged 14:25:03 <zhaochao> If you want some fixes merged into the stable branch, please add the core stable team members to the reviewer list 14:25:57 <fanzhang_> IMHO, it's better this way right now, since we don't have much time focus on the upstream community as our full-time job :-( 14:27:18 <zhaochao> agreed, I may be kind of full-time from time to time, like these two weeks, :), but cannot permit that for a longer period 14:28:21 <zhaochao> OK, do you have more on the rocky update topic, gang have added another topic about cluster supporting, we could jump to that one 14:29:27 <zhaochao> then let's move on 14:29:32 <zhaochao> #topic cluster supporting 14:29:40 <gang> yes, I am not think clearly about trove cluster, need some advises from you guys 14:30:10 <zhaochao> gang: would you tell more about your thoughts about this? 14:31:09 <zhaochao> for me, "Remove all instance API and treate instance as a single node cluster." seems not the right way, but I don't know what's your detailed plan 14:31:09 <gang> I get some thought from the email and from my production experience. 14:31:18 <gang> And I had add to the agenda 14:32:15 <gang> for production env, cluster is more import than instance. 14:32:41 <zhaochao> gang: agree with the importance of cluster in product env 14:32:54 <fanzhang_> +1 14:33:07 <gang> So, my opinion is maybe we should re-architect clusters. 14:33:17 <gang> But it seems a big change. 14:33:18 <zhaochao> however from the api level, cluster is already the first citizen as instance 14:34:05 <zhaochao> and simple instance as single-node cluster, seems not correct and not easy to achieve 14:34:18 <zhaochao> we also support replication of instances 14:34:31 <fanzhang_> As I mentioned before, we have to figure it out what has been supported in trove, and what has not. And then we can add more useful and urgent functions that user desires most 14:35:04 <fanzhang_> as for re-architect, if you got the time, well, I think maybe you can try 14:35:22 <fanzhang_> detail spec should be needed definitely 14:35:54 <gang> well, if we can split it into some smaller task, maybe I can try. 14:36:25 <gang> It just what I thought trove should looks like... 14:37:15 <zhaochao> gang: fanzhang_ yes, bp is needed. the first design is important, any good ideas about this? 14:37:49 <gang> Like replication of instances, I mean maybe it should be "transparent" to end user. 14:38:16 <zhaochao> or what are problems do you think about the current design of clusters? 14:38:56 <zhaochao> gang: ok, I may got some points from that, replications could be a kind of cluster(or strategy, maybe) 14:39:31 <fanzhang_> the end users may want to know what they're using, replication or real cluster, that really depends on them, not us. 14:41:53 <zhaochao> fanzhang_: from gang's idea, we could design something like "cluster type" -> replication/sharded/... 14:43:29 <fanzhang_> yeah, exposing cluster type to end users may be a great feature 14:43:57 <gang> fanzhang_: yes, in some case user want to know the details, this is why we discuss. 14:44:10 <zhaochao> however we still have a lot changes for this 14:44:51 <gang> zhaochao: yes, if we do this, it will be a huge work. 14:45:37 <gang> so, do you think there need a spec? 14:45:56 <zhaochao> gang: would you add the related links in https://etherpad.openstack.org/p/trove-priorities-and-specs-tracking, under "Improve cluster supporting" 14:46:03 <zhaochao> gang: yes, we need a spec 14:46:23 <gang> OK 14:46:26 <zhaochao> gang: we may need more discussions about this 14:46:33 <fanzhang_> definitely 14:46:54 <gang> zhaochao: can't agree more 14:47:37 <zhaochao> gang: if you like to push this, you could create a new pad for it, or add your thoughts under "Improve cluster supporting" of the tracking page 14:47:57 <zhaochao> everyone could help to improve this 14:48:19 <zhaochao> I think this coud be a good direction 14:48:36 <gang> zhaochao: I will, if I have some step that I would ask you and fan. 14:49:33 <fanzhang_> would love to help with that 14:50:10 <gang> fanzhang: thanks in advance :) 14:50:18 <zhaochao> gang: great, I will aslo do some investigation about this topic after I finish the global goals and Zuul v3 migration(may not take too long, one or two weeks maybe) 14:50:42 <gang> zhaochao: thanks! 14:51:24 <zhaochao> #action gang do the first draft design about improving cluster support, fanzhang and zhaochao will help on the that 14:51:28 <zhaochao> gang: no problem 14:52:34 <zhaochao> So any thing more about this topic? gang fanzhang_ 14:53:02 <gang> no 14:53:13 <fanzhang_> no 14:53:40 <fanzhang_> really tired from work and gym 🤣 14:54:02 <zhaochao> fanzhang_: good for gym 14:54:18 <zhaochao> OK, I'll go the open discussiong topic 14:54:22 <zhaochao> #topic Open discussion 14:54:40 <kumarmn_> I wanted to discuss on the Trove project updated at the Summit in Vancouver 14:54:54 <kumarmn_> I am not able to go due to a personal trip/vacation. 14:55:30 <kumarmn_> Any of you planning to go? 14:56:17 <zhaochao> kumarmn_: sorry, I'm not going 14:56:22 <fanzhang_> sorry, can not make it 14:56:28 <zhaochao> kumarmn_: I just forgot another question about the rocky-1, do we have a deadline? tomorrow? 14:56:44 <kumarmn_> Sometime this week, preferably by tomorrow. 14:56:49 <gang> me too... sorry 14:57:10 <zhaochao> kumarmn_: got that, thanks 14:57:44 <zhaochao> kumarmn_: I don't know wether any other core member will go the Vancouver summit 14:58:00 <kumarmn_> I will check with Maciej 15:00:10 <zhaochao> OK, most of the other core members are in China, I didn't hear some of them will go to the summit. fanzhang_ , do you know some of us will go? 15:00:24 <fanzhang_> no sorry 15:00:41 <zhaochao> OK, time's up 15:00:49 <zhaochao> that's all today, thanks, everyone 15:00:54 <fanzhang_> will ask songjian tomorrow 15:01:01 <zhaochao> fanzhang_: thanks 15:01:06 <zhaochao> #endmeeting