17:01:50 #startmeeting CongressTeamMeeting 17:01:51 Meeting started Tue Feb 17 17:01:50 2015 UTC and is due to finish in 60 minutes. The chair is thinrichs. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:01:52 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:01:54 The meeting name has been set to 'congressteammeeting' 17:02:17 Let's start with status updates. 17:02:19 #topic status 17:02:37 rajdeepd: since it's so late for you, want to start so you can hop off when you need to? 17:02:59 ok sure 17:03:22 i have been working on horizon data source status table 17:03:43 updated the CL based on review comments from jwh and arosen 17:04:07 sent it for review today morning india time 17:04:28 that summarizes my status update 17:05:20 Thanks. I'll try to remind them to give you another review in the next couple of days. 17:05:24 I don't see them online now. 17:05:44 sarob: want to go next? 17:05:51 shirely 17:06:02 #link http://lists.openstack.org/pipermail/openstack-dev/2015-February/056664.html 17:06:04 ok thanks 17:06:44 we want to tidy up launchpad 17:07:09 so bugs and blueprint assignments get 17:07:26 published as completed as part of the 17:07:35 of the completed milestone 17:08:08 also tagging the repo with the same milestone 17:08:45 ttx and fungi have responded to the thread 17:08:58 i will do the tasks manually this time 17:09:20 next milestone i will see if I can use ttx's script 17:09:29 any questions? 17:09:33 Hiya, Sorry i was running late. 17:09:51 arosen1: morn 17:09:55 arosen1: we're doing status updates, as usual. 17:10:11 sarob just went over our kilo2 release plans. 17:10:33 sarob: seems okay to me. 17:10:35 cool, yea i saw his email on the mailing list last week. SOunds good to me. 17:10:45 coolo 17:10:56 sarob: have you looked at how many of our kilo2 bps are slipping to kilo3? 17:11:08 hold a sec 17:11:51 #link https://launchpad.net/congress/+milestone/kilo-2 17:12:28 maybe one will be kilo-2 17:12:47 or two 17:12:58 rest are going to kilo-3 17:13:36 #link https://review.openstack.org/#/c/150514/ 17:14:01 i believe completes 17:14:05 #link https://blueprints.launchpad.net/congress/+spec/murano-driver 17:14:36 i think 17:14:41 #link https://blueprints.launchpad.net/congress/+spec/modal-operators-for-policy 17:14:46 had some code pushed 17:14:54 cant see the reference in the bp though 17:15:00 Have you heard from Kishan about the action-execution interface? 17:15:34 he was going to start work on it 17:15:49 as the murano driver was completed 17:16:05 What about Zhenzan and the policy-engine triggers? 17:16:07 i havent seen him push any code 17:16:50 hi. I'm working on murano-driver. I still have some revision I plan to submit for review. 17:16:50 nothing since last week 17:17:09 kishan around? 17:17:35 zhenzan? 17:17:49 stevenldt: Great! Those drivers take a number of revisions before they're stable, typically. 17:18:08 stevenldt: want to do your status update? 17:18:26 i will follow up with those two 17:18:29 im done 17:18:29 sure 17:19:12 I'm working on some revision to extend the datasource tables to sync up with Murano 17:19:23 sarob: sorry—assumed you were done, but I should have asked. 17:19:47 #action sarob follow up with kishan action-execution and zhenzan policy engine triggers 17:20:03 I may have the code ready for review this week or early next since this week is a short week for me 17:20:04 thinrichs: np! 17:21:14 stevenldt: sounds good. Let us know if you need anything from us. 17:21:33 I have a question regarding the changes on the datasource query, at least on the cli 17:21:50 stevenldt: shoot 17:21:53 #action sarob will release kilo-2 milestone; launchpad milestone kilo-2 release and github tagging repo with 2015.1.0b2 17:22:37 we now have to create the datasource before use, and to read the datasource table, we use the uuid instead of the name of the datasource 17:22:50 is that what we're going forward with? 17:23:06 stevenldt: yup. 17:23:16 From the CLI, we shouldn't need to use the datasource uuid. 17:23:20 This allows us to move to a multitenant api as well. 17:23:20 That's a step backwards. 17:23:34 thinrichs: right you don't need to use the uuid on the CLI 17:23:38 Ok. Thanks. I may need to review my part to adapt to that. 17:23:58 I think there was one bug there where you did need to use the cli but here it is: https://review.openstack.org/#/c/154993/ 17:24:09 thats the fix ^ 17:24:44 Keeping it usable by humans is a ++. 17:25:08 yup you can keep using the CLI by name or uuid. 17:25:16 stevenldt: anything else? 17:25:20 Great. That's all from me. 17:25:43 cloudtoa_: want to do a status report? 17:26:06 I need 36 hours in a day. 17:26:53 I'll have something submitted in the next couple of days for the control bus part. 17:27:25 Assuming I have time, I'll have the first pass at the blueprint for the table-service submitted. 17:27:32 By end of week. 17:28:16 The kilo3 deadline is mid-March. 17:28:39 After that we stop adding new features, for the most part. 17:28:46 And we focus on stabilizing. 17:29:01 Before the official kilo release. 17:29:26 cloudtoa_: sounds like you'll have the control bus work done by kilo3. 17:29:31 yes 17:29:33 What do you think about the table-service work? 17:30:50 What do you mean? You mean the work that arosen is doing? 17:31:41 cloudtoa_: you just said you'd make a first pass at the blueprint for the table-service work. 17:31:46 That's what I'm asking about. 17:32:26 I was thinking of splitting the datasource driver into different pieces. It's probably too much to squeeze into an IRC session. 17:33:08 OK. We'll wait for your blueprint/spec then. 17:33:17 cloudtoa_: anything else from you? 17:33:27 No, that is all. 17:33:31 Thanks! 17:33:43 arosen1: want to tell us what you've been up to? 17:35:17 Sure 17:35:41 So, we just merged the patch that allows datasources to be configurable via the api on wednesday last week. 17:36:06 Now we no longer have the datasource.conf file and all the datasource information is stored in the database instead. 17:36:51 In addition, several changes were made to the api code to use uuid's in the URLs instead of names so that our api can eventually work with multiple tenants. 17:37:29 I've also been revamping our CI a little and now the congress ci also validates patches against the python-congressclient. 17:37:39 That's it unless someone wants to discuss. 17:38:13 datasource.conf being gone... 17:38:20 are we storing like a JSON blob in the database? 17:38:29 cloudtoa_: yup 17:38:42 there is a new api call /v1/drivers//config 17:38:45 Is there a way to update that without deleting the row? 17:39:13 which tells you want you need to pass in to congress to register a specific datasource drivers (since they all have different connection requirements) 17:39:40 ah, the'/config' part was missing in the git commit. That was confusing me a bit. 17:39:41 Radu-: right now we only support POST/DELETE/GET on it. I didn't implement update 17:40:20 Radu-: yea, check out the devstack changes I made as well which shows how to use it via the cli 17:40:26 it doesn't show getting the config though 17:41:33 Hello, everyone. It is the first time for me to enter this meeting. But Jim Xu has exchanged some ideas about congress UI. I hope I can do some contributions. 17:41:41 and it seems that you have to specific what values can be configured from the driver as well now. 17:42:23 Yali_: Nice to have you join us! After we finish up discussing these latest changes from arosen, we'll ask you to talk a bit about your interests. 17:42:28 Radu-: yup 17:43:17 is that just a matter of updating the drivers get_datasource_info result[config] variable? 17:44:44 Sorry If Im repeating anything that was discussed when I had left the channel for a few minutes. Didn't think I would be able to stay any longer 17:44:50 Radu-: https://github.com/stackforge/congress/blob/master/congress/datasources/plexxi_driver.py#L83 17:44:58 Radu-: yea lets talk about it in #congress after 17:45:08 Ok thanks 17:45:24 arosen1: Which APIs changed to require UUIDs instead of names? 17:46:28 thinrichs: the data-sources one 17:46:42 since now an instance of a data-srouce is a uuid not a name 17:46:50 in the url 17:47:16 * arosen1 but the client hides all of that for you if you want to use the name 17:47:44 Just the datasource then? 17:48:14 correct 17:48:49 OK. I think as a general rule we want to make the API friendly as well, in case someone isn't using the CLI. 17:49:23 I should give a quick status update. 17:49:59 I've been working on the problem of delegation recently: how do we carve off a piece of the policy the user has given us and give it to a domain-specific policy engine, eg. a VM-placement policy engine? 17:50:24 You can find details if you look at the openstack-dev mailing list for messages with [Congress][Delegation] in the subject. 17:50:41 Anyone else have a status update they want to give, before we hear from Yali. 17:50:43 ? 17:51:41 Yali_: could you tell us a bit about yourself and why you're interested in Congress? 17:52:02 sure 17:53:21 i an a newer for congress. but i think it is very important for the whole cloud. 17:53:44 to make the cloud compliance. 17:54:42 Could you tell us your name and where you work, so we know you outside of IRC? 17:55:43 Ok. My name is Yali Zhang, from China. And i am working in Huawei Company, same with Jim Xu. 17:56:37 Great! Is there some part of Congress you're especially interested in? Policy engine, datasources, UI? 17:57:58 Yes, I am interested in UI nowadays. 17:58:24 Glad to have you. Let us know if you need pointers to get started. Many of us hang out in #congress in IRC all day. 17:58:30 2 minutes remaining. 17:58:38 Any topics to broach quickly? 17:58:42 #topic open discussion 17:59:45 im working on getting a design session room for at 60+ people for at three hour sessions 17:59:48 now I have some ideas about UI, and hope to share with you. 18:00:07 sarob: nice! 18:00:24 If the next summit is anything like the last we'll need it. 18:00:28 yup 18:00:31 Yali_: we're looking forward to it! 18:00:37 And we're out of time. 18:00:37 ill update on incubation next time 18:00:43 Also voting for openstack summit talks are now open. 18:00:49 We can continue on #congress if anyone wants. 18:00:55 Yes—everyone get out and vote! 18:00:58 Here's a list of the congress talks if you guys want to vote for all of them: http://blog.aaronorosen.com/congress-liberty-summit-talks/ 18:01:10 #endmeeting