18:59:08 #startmeeting Keystone Team Meeting 18:59:09 Meeting started Tue Nov 8 18:59:08 2011 UTC. The chair is jsavak. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:59:11 Useful Commands: #action #agreed #help #info #idea #link #topic. 18:59:19 Agenda: http://wiki.openstack.org/Meetings/KeystoneMeeting 18:59:54 Hi! I'm here. FYI 19:00:15 Hi Ziad! 19:00:18 Who else is here? 19:00:44 o/ 19:00:55 Hey Joseph - good talking with you yesterday 19:02:18 #topic Roadmap for Essex - status on blueprints 19:02:38 jsavak: yeah, definitely! 19:02:53 jsavak: doc should be pretty reasonable now - for the BP 19:03:43 Yeah - doc looks good. I was looking at keystone.openstack.org. The only thing that may need beefing up is the curl examples, but we can come back to those. 19:03:48 I'm going through the blueprints for e-1 right now (listed on http://wiki.openstack.org/releasestatus/) and updating the blueprints as necessary. If it's merged, I've been saying that the blueprint is implemented. Is that the right status? 19:04:49 * heckj isn't entirely sure of best practice re: blueprints 19:05:05 Isn't the status of the BP supposed to get updated by Gerrit if the BP is referenced in the comments? 19:05:28 in horizon, we have some general blueprints that stay open and aren't assigned to milestones, others that are assigned and closed as they get completed. 19:05:33 maybe these weren't then - they've been showing good progress even though the code was merged 19:05:36 zns: yes 19:05:44 I think putting a BP# comment in there is what we need to do... 19:05:56 zns: hi 19:06:04 heckj: You know how to trigger that? Is it BP#? 19:06:22 zns: auto status changes in blueprints isn't implemented yet, but is a todo 19:06:23 "blueprint [name of blueprint here]" is all I've needed in the past 19:06:39 damn - that must have just been bug then - I thought it was updating 19:06:46 right now it links back to the change if you do that 19:06:50 but doesn't change the status 19:07:20 jeblair: wait - where did the updates for https://blueprints.launchpad.net/keystone/+spec/keystone-documentation come in? 19:08:04 jeblair: hi 19:08:06 yeah, those are automatic based on saying something like "blueprint keystone-documentation" 19:08:39 i think zns was talking about updating the actual "status" field in the bp 19:08:41 one consideration though for that keystone-documentation blueprint is that the files in openstack-manuals (which publish to docs.openstack.org) are not yet updated 19:08:55 jeblair: OK, so status still needs to be managed manually for now. np 19:08:58 annegentle: yeah, not yet complete - just dev docs 19:09:04 which we've talked about doing but don't do yet. 19:09:06 zns: yes 19:09:18 I met with Joe and Theresa and we still need an assignee for those updates 19:10:07 zns: speaking of which, I'm still waiting on answers to questions from you, dolph, or yogi. Posted to mailing list after last keystone meeting I was in - never got a response from you guys. 19:11:11 heckj - was it the "why does it" email? 19:12:18 I saw that vish and jesse responded so I wasn't sure if there were any outstanding questions... 19:12:19 jsavak: yeah - the questions you said that I should send to the mailing list last meeting 19:12:33 Several, I'm afraid 19:13:00 ping dolphm_ are you here? 19:13:26 so there is a ci team meeting scheduled here for this time slot.... 19:13:28 they answered the credentials question, but that led to "how do you get keystone to work with EC2" - it really needs whomever is forward thinking on this project to answer. 19:13:41 heckj: sorry, I missed that one. I just flagged it and will spend time on it this evening. 19:13:49 jeblair - whoops 19:14:06 heckj - can you send me your outstanding questions? I'll get them to dolph & yogi 19:14:33 one the way 19:14:54 heckj - we're also going to spend more time in irc & responding to mailing lists. We've been bad about this in the past 19:15:20 Topic: Open Discussion re: keystone 19:15:29 #topic open discussion re: keystone 19:15:41 heckj: We've done some work on the EC2 stuff. Yogi has been working on it. I'll try to put more meet on the answers this evening. 19:15:44 jsavak: yeah, i'm here, i've been following along 19:16:12 zns: cool, thank you 19:16:19 dolphm: can you work with heckj on the questions he has? 19:16:27 ins, jsavak: need that to wrap from narrative around the docs update 19:16:38 jsavak: not sure how we closed the previous topic. Are we on track for Diablo patch and E1? 19:16:49 jsavak: i don't know the answers to the ones remaining, but i can help him research :) they're good questions 19:17:12 zns: diablo patch, yes 19:17:25 zns: we are on track. All of the blueprints mentioned on http://wiki.openstack.org/releasestatus/ have been merged. 19:17:42 er - metioned for e1 have been merged 19:17:49 jsavak, dolphm_: thanks. Cool! 19:18:24 zns: I'd like your review for the Keystone conceptual diagram at https://review.openstack.org/#change,1033 before pushing it through 19:18:33 zns: we need to publicize your RBAC prototype. Can you send out a mailing-list email on it? 19:18:37 <_0x44> sandywalsh: You around? Look at #openstack-dev 19:19:17 jsavak: sure. I wanted to test it with devstack, but maybe we shouldn't wait. I'll send it out today. 19:19:32 zns: thanks! 19:19:52 annegentle: OK. Will do. 19:20:13 any other keystone business or questions? 19:20:24 jsavak: yeah... review 1068 19:20:25 Also are all Keystone's API updates targeted to 2.0? Does Keystone need a RFC period on its API docs? I'm asking since Compute and Netconn (Quantum) and Image (Glance) need RFC periods it looks like. 19:20:50 So I'm collecting API draft spec needs from teams. 19:20:51 we need to notify the community before that merges -- it's got a couple of backwards-incompatible changes that WILL break *every* consumer of keystone 19:20:55 annegentle: there are no API updates as far as I know…. 19:21:28 zns: ok so we'll be on API 2.0 from now until April (Essex) for certain? 19:21:41 annegentle: yes. 19:21:48 zns: ok thanks 19:22:05 clarification - the only thing we will add will be non-breaking extensions... 19:22:59 zns: and all - the key I think is, do you need community feedback through a request for comments period while you update the api? Maybe I'm asking the wrong question... 19:23:01 dolphm: is 1068 close to submission? 19:23:15 1068 - we need to make sure that the instructions are clear and the experience is smooth. I don't think we're there yet. 19:23:17 jsavak: sort of 19:23:23 jsavak: i think it should go in after E1 19:23:54 dolphm: ok 19:24:06 annegentle: Yes, but we don't have an API change scheduled for Essex. We may have some proposals for a new version at the F summit. 19:24:12 zns: ok 19:24:17 jsavak: talking with yogi about it now 19:24:29 dolphm_: agreed. 19:24:30 dolphm: ok, thanks. 19:25:22 jsavak: what I'd like to see is the RCB guys test it out and tell us it meets their needs/expectations. 19:25:43 zns: test out 1068, or e1 overall? 19:26:06 or both 19:26:13 jsavak: i think he means the rbc prototype 19:26:20 1068. They would be a good bar to test any release against, but especially a hairy one like 1068. 19:26:40 zns: ok. 19:26:54 Dolph: let me know when 1068 is looking better and I'll get jesse to pull & review 19:27:03 dolphm_: no - not the prototype…. 1068. 19:27:19 zns: all of our commits go through RCB CI already, afaik, but the coverage is poor 19:28:26 dolphm_: I'm thinking about the production deployments, not just the CI server. 19:29:20 ok - any other keystone questions or issues? 19:30:21 Ok - thanks for your time. I'll post the minutes on the wiki. 19:30:30 #endmeeting