14:01:12 <isviridov> #startmeeting magnetodb 14:01:17 <openstack> Meeting started Thu Mar 26 14:01:12 2015 UTC and is due to finish in 60 minutes. The chair is isviridov. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:18 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:20 <openstack> The meeting name has been set to 'magnetodb' 14:01:24 <ajayaa> Hi isviridov. 14:01:25 <isviridov> Who is here today? 14:01:29 <isviridov> o/ 14:01:30 <ajayaa> o/ 14:01:37 <charlesw> o/ 14:02:20 <isviridov> Ok, today agenda #link https://wiki.openstack.org/wiki/MagnetoDB/WeeklyMeetingAgenda#26_Mar.2C_2015.2C_14:00_UTC 14:02:34 <isviridov> And the first short topic 14:02:38 <isviridov> #topic Go through action items isviridov 14:02:50 <isviridov> #link http://eavesdrop.openstack.org/meetings/magnetodb/2015/magnetodb.2015-03-12-14.01.html 14:03:14 <isviridov> There was only one, and I can report that meeting schedule has been updated 14:03:22 <isviridov> Now it is every second week 14:03:27 <dukhlov> o/ 14:03:38 <ajayaa> I saw the logs of previous meeting. 14:03:45 <ajayaa> fine with that btw. 14:03:51 <isviridov> #link https://wiki.openstack.org/wiki/Meetings#MagnetoDB_Team_meeting 14:04:16 <isviridov> Hello ominakov 14:04:26 <isviridov> ajayaa great! 14:04:33 <ominakov> hello guys 14:04:35 <isviridov> No other action items 14:04:38 <isviridov> Next topic 14:04:48 <isviridov> #topic Round of status isviridov 14:04:58 <isviridov> Who want to be first? 14:05:17 <charlesw> I'll go 14:05:49 <isviridov> ominakov welcome back 14:06:44 <charlesw> Since last meeting, we merged some patches: refactoring notification, adding health check body details, standalone health check metrics service. 14:07:50 <charlesw> Dima helped a lot. Truely a team effort. 14:08:08 <isviridov> Great to hear! 14:08:26 <isviridov> I can share my status 14:09:02 <isviridov> Actually I didn't spend a lot of time to project last weeks. 14:09:34 <isviridov> But I've ordered launchpad and released kilo-3 milestone 14:09:43 <isviridov> Also send short anonce in ML 14:09:49 <isviridov> #link https://launchpad.net/magnetodb/kilo/kilo-3 14:09:52 <isviridov> Over 14:10:45 <isviridov> dukhlov ajayaa ominakov any updates from your side? 14:11:11 <dukhlov> isviridov, any plans about openstack incubation, 14:11:12 <dukhlov> ? 14:11:22 <charlesw> Maybe you can share the vision going forward. 14:11:53 <isviridov> dukhlov charlesw sure 14:13:31 <isviridov> The latest update from TC I know was an annoncement that new big tent reform is applyed and now any project can apply for status as part of openstack 14:14:27 <isviridov> #link http://www.openstack.org/blog/2015/02/tc-update-project-reform-progress/ 14:15:42 <ajayaa> isviridov, We are working on a mdb backend for Openstack services. We have ported Keystone to use MagnetoDB as the datastore. 14:16:18 <isviridov> So, now we have a clear rules how to start. However I have not found any recomened way how to do it. 14:16:47 <ajayaa> We are trying to write a scalable and fault tolerant db layer for Openstack services. 14:17:03 <charlesw> ajayaa, sounds very cool, any details, links? 14:17:07 <ajayaa> isviridov, Am I interrupting you? 14:17:19 <ajayaa> http://ajayaa.github.io/distributed-db.html 14:17:27 <isviridov> Also personally it is not my first priority and I can't spend 100% time on it, so any help or ever initiative to drive this process is very appresiated 14:17:32 <ajayaa> here is a demo of Keystone running with mdb as backend. 14:17:39 <isviridov> ajayaa it is ok. N.p. 14:18:07 <isviridov> The showrt answer, I 14:18:50 <isviridov> 'll try to start preparation the documents/wiki pages next week according to https://wiki.openstack.org/wiki/Governance/NewProjectTeams 14:18:54 <ajayaa> Here is a spec for the same. https://review.openstack.org/#/c/148521/ 14:19:02 <isviridov> If anyone would like to joing very welcome 14:19:16 <isviridov> charlesw that is about incubation 14:19:41 * isviridov checking links from ajayaa 14:21:08 <ajayaa> The keystone spec will get accepted once Keystone people start reviewing specs for L. 14:21:32 <isviridov> ajayaa looks really cool, have you just removed keystone integration? 14:21:46 <ajayaa> isviridov, yes. needed to. 14:22:15 <ajayaa> otherwise we would be stuck in a loop. :) 14:23:02 <isviridov> Unexpected usage actually 14:23:37 <ajayaa> actually, we have plans to move all of OS components to mdb in the future. 14:23:52 <isviridov> #What are your plans, about it? 14:23:56 <isviridov> I see 14:24:00 <ajayaa> The real problem is OS apis were written initially assuming relational db as a backend. 14:24:21 <ajayaa> But we are trying our best to support all the apis. 14:24:39 <ajayaa> We have started to move glance as well. 14:24:48 <ajayaa> Nova is the real challenge. 14:25:30 <isviridov> ajayaa what is feedback? 14:25:42 <ajayaa> Also I wanted to let you people know that dynamodb api is a bit limited. 14:25:50 <isviridov> I mena from glance and other you have tried already 14:26:38 <ajayaa> We are designing schema for glance now. But we have a working Keystone as I showed in the demo. 14:26:51 <dukhlov> ajayaa, looks like mdb is not really good backend for glance, it needs to store huge images 14:26:54 * isviridov for protocol #link https://review.openstack.org/#/c/148521/2 14:27:12 <ajayaa> dukhlov, it does not store images on its own. 14:27:25 <dukhlov> only metadata? 14:27:35 <ajayaa> dukhlov, yes. 14:27:44 <dukhlov> ah, ok 14:27:47 <ajayaa> The backend is swift or any object store. 14:29:08 <ajayaa> The fact that mdb allows at maximum two columns per key is quite limiting btw. 14:29:45 <ajayaa> especially when you are denormalizing data and storing it into a single table. 14:30:46 <dukhlov> we can change it now I guess 14:30:51 <ajayaa> But C* allows that. Can we by any chance allow more than two columns per key? 14:31:34 <ajayaa> I would propose a spec for that if everyone is in favour of this? 14:31:38 <dukhlov> now yes, after index LSI redesign I think we can extend our functionality 14:31:59 <isviridov> ajayaa I think we can, if we don't see performance impact 14:32:18 <ajayaa> isviridov, dukhlov, C* already allows that. 14:32:29 <dukhlov> and still support DDB API 14:32:51 <ajayaa> We will just be a superset of ddb api. 14:33:47 <dukhlov> ok, you can start to write spec 14:34:24 <ajayaa> dukhlov, cool. 14:34:25 <dukhlov> I can't guarantee now that we don't have anty problem there 14:34:49 <ajayaa> anty? 14:34:50 <dukhlov> but I think we will figure out this during reviewing spec 14:34:57 <dukhlov> any 14:35:25 <isviridov> ajayaa have you shared your results with community? 14:35:30 <ajayaa> I will write a spec. 14:35:52 <ajayaa> isviridov, not yet. not the results. Just the spec with Keystone community. 14:36:40 <isviridov> I think it is important, to spread the info. NoSQL and Java always causes huge discussions 14:39:01 <isviridov> Looks like no other status. Next topic? 14:40:17 <ajayaa> isviridov, I am doing that. 14:40:29 <ajayaa> I am also attending the glance meeting now. 14:40:40 <ajayaa> time clash. 14:41:13 * isviridov also joined 14:42:03 <isviridov> #topic Open discussion isviridov 14:42:09 <isviridov> The last topic 14:44:20 <isviridov> dukhlov charlesw do you think we a done for today? 14:46:56 <isviridov> Thank you guys for attending today meeting 14:47:03 <isviridov> #endmeeting