21:02:31 #startmeeting project 21:02:31 Meeting started Tue Oct 23 21:02:31 2012 UTC. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:32 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:33 The meeting name has been set to 'project' 21:02:35 o/ 21:02:38 Agenda @ http://wiki.openstack.org/Meetings/ProjectMeeting 21:02:54 #topic Design Summit feedback 21:03:07 While it's hot, any comment on the OpenStack Summit in general, and the Design Summit track in particular ? 21:03:29 the content/audience match was off for me 21:03:36 Did the format work OK ? (4 days with 4 parallel topics + 1 unconference, running parallel to rest of Summit) 21:03:45 i enjoyed being there, but i was notincg a scaling issue with the design meetings 21:04:09 I think paralell introduced some challenges in the design sessions 21:04:14 MarkAtwood: as in too many people in design summit sessions ? 21:04:14 by day 4 of design summit sessions i was fried. 21:04:15 I missed being able to attend more of the "use case"-type talks, but otherwise didn't mind the format. 21:04:25 My voice was fried on day 1 21:04:31 yeah, i felt like there were a lot of "conference" people in the "summit" session. Also, after 3 days of sessions, I was fried. 21:04:32 more user feedback during sessions == goodness, but less brainstorming/design that didn't go off the rails with lots of relatively uninformed voices 21:04:33 yeah, i didn't get to any of the non-design sessions, which would have been nice 21:04:56 4 days was pretty brutal for nova 21:05:00 But having people asking how object store in cinder works in design sessions was wasteful 21:05:10 vishy: could you have done with less ? 21:05:33 some of the nova sessions could have been vetted more and then not done at all 21:05:33 * vishy didn't notice the issue that danwent and notmyname are mentioning 21:05:39 though I agree on the occasional audience mismatch, I also found I ended up talking to people who otherwise wouldn't have been in design sessions and that's (occasionally) good for growing the base. 21:05:42 maybe they flocked to quantum and swift :) 21:05:47 :-) 21:05:51 so we probably could have done 3 days of good sessions 21:05:56 quantum had a huge number of quiet people in the room... 21:06:04 zomg networkz 21:06:07 the trick is that there is no good way of refining attendance, short of making the design summit invite-only..; which sends the wrong message 21:06:13 quantum was attrociously packed 21:06:18 yeah 21:06:27 ttx: I think splitting helped with some of that 21:06:29 the user sessions were over cap too 21:06:48 so plan B (try to attract business types in other rooms) kinda worked 21:06:48 it may be worth social engeineering, put in the conference schedule, the guids, and in signs on the doors to the sessions 21:06:54 well, i think its really that there's the overlap between the conference and the summit, and conference people decided to hang out in sessions, even though there was no intent to contribute code, and no experience using openstack. 21:07:04 danwent: ++++1 21:07:16 "this is an advanced session, not a training session, please contribute if you are up to speed" 21:07:27 rewritten for grace and diplomacy, of course 21:07:29 if this happened for jgriffith but not vishy, perhaps its more for the newer projects where people are more curious 21:07:34 MarkAtwood: there were pretty obvious signs that they were entering development zone 21:07:37 since "ci" topics were on a separate track, it meant that we often didn't get people we would have liked. having more PTLs to talk about testr and distribution support would have been esuful. 21:07:42 but maybe we can separate the events more 21:07:50 ttx: that is my thinking 21:07:55 Next time, 4th floor ! 21:08:01 LOL 21:08:04 Or solve a riddle to get there 21:08:05 basement! 21:08:09 No escalator! 21:08:09 with a scary entrance 21:08:19 haha… time separation > space separation 21:08:26 Even just marking how "advanced" a particular session is might be helpful 21:08:31 time separation won't work. They will come 21:08:56 it was better before we combined days 21:08:56 require foundation membership and/or signed contributoir agreement? 21:09:00 Regardless, it was workable but I was done by Thursday for sure 21:09:02 (at least from my perspective) 21:09:06 how about a terminal where you have to do a series of git commands to unlock the door 21:09:11 vishy: ++ 21:09:14 vishy: Nice!!! 21:09:14 :) 21:09:16 vishy, :P 21:09:22 vishy: a lot of devs would fail that. :( 21:09:24 I say we vote! 21:09:51 ok, well if you've a good suggestion, please send it my way 21:09:52 can you clone the nova repo? You're in 21:10:11 or a simple python task ... 21:10:11 since format on the next one should be decided pretty quickly 21:10:22 by default, the same format will be done again 21:10:32 though we can discuss separation / special badges 21:10:39 * jgriffith will start training for it 21:10:47 and/or reducing days / parallel tracks 21:10:54 #topic Grizzly release schedule 21:11:04 We looked into the proposed schedule at the Design Summit 21:11:09 Here is the final proposal: 21:11:14 #link http://wiki.openstack.org/GrizzlyReleaseSchedule 21:11:18 (no change since Friday) 21:11:23 * gabrielhurley likes special badges 21:11:24 err Thursday 21:11:41 Business summary: It is the same as the Folsom schedule, with one additional week in the second milestone to account for the Christmas holidays. 21:11:53 Any comment before we officialize it ? 21:12:55 I guess not. 21:13:07 #agreed http://wiki.openstack.org/GrizzlyReleaseSchedule 21:13:17 #topic Oslo status 21:13:24 russellb: hey 21:13:30 hey 21:13:35 So I added Oslo (openstack-common library name) as a regular status report in this meeting 21:13:41 sounds good. 21:13:53 russellb: does markmc want to track the blueprint work at launchpad.net/openstack-common or launchpad.net/oslo ? 21:13:53 so out of the summit, moving forward with libraryification is the #1 priority 21:14:11 need to create a grizzly series wherever you prefer 21:14:12 good question. we already have a lot of history in openstack-common ... can be it be moved? 21:14:37 it.. can. Might be shorter to rename project though 21:14:55 #action ttx to see with markmc to move openstack-common to oslo 21:15:12 #action ttx to create grizzly series wherever needed 21:15:22 (and milestones) 21:15:24 looks like oslo-cfg is the targeted first library release 21:15:33 The idea is to come up, in the next two weeks, with a good roadmap for Grizzly objectives 21:15:54 Get people to file blueprints, set series goal to "grizzly" where appropriate... prioritize them and target to a given milestone 21:16:02 k, markmc will be back next week, i don't think it will be a problem to work through the roadmap before 2 weeks from now based on summit output 21:16:05 we might need to clarify the versioning question first 21:16:23 #action ttx to discuss oslo versioning with mordred and markmc 21:16:49 cool, don't think I have anything else to report 21:16:52 russellb: Anything else you wanted to mention ? 21:16:58 Questions on Oslo ? 21:17:23 yes 21:17:34 bcwaldon: go for it 21:17:35 I'd like to hear an explanation of the scope of Oslo 21:17:44 and what code should live there 21:18:04 code useful for multiple openstack projects, but not necessarily generally useful outside of openstack 21:18:14 russellb: thats still rather generic, no? 21:18:19 it is, indeed. 21:18:25 but Oslo will not be one library 21:18:31 russellb: ok, if thats then intention, then ok 21:18:35 the plan is to have multiple libraries released under the oslo namespace 21:18:38 http://wiki.openstack.org/CommonLibrary has a good set of rules of thumb 21:18:47 russellb: and I'd love to hear when we can consume Oslo as an importable library rather than carrying it everywhere 21:19:00 i might just be out of ze loop 21:19:03 asap basically, but it's one API at a time 21:19:09 bcwaldon: +1 - I know it's the top priority 21:19:11 so oslo-cfg will be the first library 21:19:22 ok, carry on then 21:19:51 Any other question ? 21:19:56 understand the managed copy-paste thing is a pain, but hopefully it's better overall than unmanaged copying around 21:20:12 russellb: yes, but only slightly 21:20:16 markmc had a good list of library candidates at the summit session, he should be able to publicize it more clearly 21:20:17 :) 21:20:23 #topic Keystone status 21:20:24 yup 21:20:28 heckj: o/ 21:20:35 #link https://blueprints.launchpad.net/keystone/grizzly 21:20:43 #action ttx to create milestones according to Release Schedule 21:20:48 collating blueprints from the summit and starting to work out who's going to do what 21:20:54 heckj: you should already be able to target the ones that will land in grizzly-1 21:21:05 heckj: same as oslo, the idea is to come up with a good roadmap in the next two weeks 21:21:13 ttx: agreed 21:21:38 Set the series goal for all the relevant ones to "grizzly" 21:21:47 You can also clean up https://blueprints.launchpad.net/keystone by setting some of them obsolete/superseded 21:22:07 Note that there is one that was "proposed" for Grizzly, that you may want to review @ https://blueprints.launchpad.net/keystone/grizzly/+setgoals 21:22:08 started, but not complete in both spaces, as well as generally prioritizing them all 21:22:19 heckj: Cool. Anything else ? 21:22:44 Did your sessions go well ? 21:22:49 great feedback at the summit, looking to take the general interest into solid implementation over the next couple of months 21:22:57 Questions about Keystone ? 21:23:21 #topic Swift status 21:23:24 o/ 21:23:25 notmyname: hi! 21:23:32 Any idea what your next version should be called ? 1.8.0 ? 1.7.6 ? 21:23:43 I'd like to create the first Grizzly milestone for you so that stuff can be targeted to it -- maybe we can pick a name and rename it later if you change your mind ? 21:23:51 we're still tracking for 1.7.5 on trunk 21:24:11 maybe I should just call it 1.7.5 and rename it if needed later ? 21:24:15 ya 21:24:23 so far, I expect this one to be 1.7.4 21:24:25 err 21:24:27 1.7.5 21:24:32 #action ttx to create 1.7.5 milestone in swift/grizzly 21:24:35 Also you have one "proposed for Grizzly" blueprint in https://blueprints.launchpad.net/swift/grizzly/+setgoals that you may want to review 21:24:58 notmyname: anything else ? How were the Swift sessions at the Design Summit ? 21:25:28 the sessions were ok. lot's on interested but uninformed people at the design sessions. but we had some good discussions at times 21:26:09 overall I'm excited about what things will be added in grizzly 21:26:17 Re: uninformed people: be sure to mention that in the official survey that should come out soon -- will give us more leverage to negociate more separation 21:26:28 Questions on Swift ? 21:27:06 #topic Glance status 21:27:11 bcwaldon: o/ 21:27:13 hey hey 21:27:14 #link https://blueprints.launchpad.net/glance/grizzly 21:27:18 Looks well advanced already... 21:27:25 why thank you 21:27:28 You might want to parse https://blueprints.launchpad.net/glance and mark obsolete/superseded accordingly 21:27:35 yep 21:27:49 havent been able to do the first pass of tactical planning after the summit yet 21:28:09 and try to come up with a reasonable grizzly plan two weeks from now 21:28:22 does that sound doable ? 21:28:31 set the bar low 21:28:33 fantastic 21:28:47 Well, I'll skip next week's meeting, so ... ;) 21:28:49 bcwaldon: Anything else ? 21:29:01 I did want to thank all those that attended the Glance summit sessions 21:29:08 they were very targeted 21:29:14 we had all the discussions we needed to 21:29:17 I missed them all, hope you wren't alone in those 21:29:23 not at all 21:29:44 not too much noise / people without a clue ? 21:29:54 nope, Glance is rather small and unintersting 21:30:04 Looks like it's only Swift/Quantum that were invaded 21:30:14 openstack-common and Process were alright 21:30:29 Questions on Glance ? 21:30:33 Glance only had 3 sessions, so we didnt give the masses much of a chance 21:30:45 Pick boring titles, that's the key 21:30:52 you've figured me out! 21:30:57 "Stable branch maintenance" 21:30:59 Image Workers 21:31:05 bcwaldon: lol 21:31:09 that was for you, markwash 21:31:17 ...moving on 21:31:24 #topic Quantum status 21:31:29 danwent: hey 21:31:35 #link https://blueprints.launchpad.net/quantum/grizzly 21:31:50 hey 21:31:59 Don't forget to set priorities when you add a blueprint to the "grizzly" series goal 21:32:06 at team meeting yesterday we went over all key blueprints that need to be filed for g-1 21:32:18 Same as others, now is a good time to go through https://blueprints.launchpad.net/quantum and mark obsolete/superseded accordingly 21:32:20 but i'd say about 2/3 have not yet been filed 21:32:33 A number of blueprints have been "proposed" for grizzly, so you might want to review them at: https://blueprints.launchpad.net/quantum/grizzly/+setgoals 21:32:38 yup, makes sense 21:32:46 8 in there 21:33:14 So yeah, start with G1 objectives, and try to come up with a longer grizzly roadmap in two weeks time 21:33:23 danwent: Anything else ? 21:33:25 nope 21:33:35 Questions on Quantum ? 21:34:29 Even if we don't have a meeting next week, expect me to push for blueprints by IRC pings... just no need to have another meeting to say "one more week!" 21:34:38 #topic Cinder status 21:34:41 jgriffith: o/ 21:34:50 #link https://blueprints.launchpad.net/cinder/grizzly 21:34:50 hey there 21:34:57 Nothing in there yet :) 21:35:00 I have 2 weeks to clean this up :) 21:35:03 There are two proposed @ https://blueprints.launchpad.net/cinder/grizzly/+setgoals 21:35:14 Looks like you can also parse https://blueprints.launchpad.net/cinder for good grizzly candidates 21:35:25 and obsolete/supersede the others 21:35:36 How dod the Cinder sessions go ? 21:35:43 did* 21:35:44 Yep, we're in the process of creating session outcomes to bp's 21:35:56 Really well for the first Cinder summit (IMO) 21:36:19 some sessions had the looky-lou problem 21:36:21 but all in all 21:36:24 Got enough time to discuss what you needed ? 21:36:49 yes, I think any more time would've just caused more trouble 21:36:53 jgriffith: Anything else ? 21:36:56 Nope 21:36:58 Questions on Cinder ? 21:37:27 #topic Nova status 21:37:31 vishy: o/ 21:37:31 cinder rooms could have been larger :) 21:37:39 hi 21:37:41 #link https://blueprints.launchpad.net/nova/grizzly 21:38:03 Also 7 proposed at https://blueprints.launchpad.net/nova/grizzly/+setgoals 21:38:10 there will be a lot more in there 21:38:15 and loads of potential cleanup in https://blueprints.launchpad.net/nova as always :) 21:38:31 I'm currently going through session notes and trying to crystalize them into decisions and blueprints 21:38:38 probably will take me until next week 21:38:38 vishy: is two weeks reasonable to come up with a realatively sane grizzly roadmap ? 21:39:47 sure 21:39:53 two weeks seems reasonable 21:39:55 vishy: you mentioned you could have done with less Nova at the summit ? 21:40:04 well i just was fried by the end 21:40:04 what would have been the good length ? 3 days ? 21:40:11 I think 3 days for nova 21:40:18 and a day off to peruse other tracks or sleep 21:40:21 would be good :) 21:40:34 then have other connected topics on the 4th day (process/qa/common etc) 21:40:50 * ttx tries to internalize all feedback 21:41:00 vishy: Anything else ? 21:41:14 nothing for me 21:41:24 vishfacts: Vish can attend all sessions of the Design Summit. 21:41:33 Questions on Nova ? 21:41:56 vish doesn't attend the design summit, the design summit attends vish. 21:42:15 #topic Horizon status 21:42:32 gabrielhurley: I didn't follow, were you Dope or Stack ? 21:42:37 I was Terrence Dope 21:42:57 http://dopenstack.com 21:42:58 you'll always be Terrence Dope now 21:43:04 #link https://blueprints.launchpad.net/horizon/grizzly 21:43:13 Looks pretty good to me already :) 21:43:18 yeah, I've been grooming it 21:43:26 You might want to parse https://blueprints.launchpad.net/horizon and mark obsolete/superseded the ones that are polluting the view, if any 21:43:29 I opened a whole slew of new blueprints after the sumit 21:43:31 rather than let them pile up 21:44:10 they're actually more-or-less all correct 21:44:17 and should be targeted to either G or H 21:44:26 gabrielhurley: did you end up having enough, or too much time on the Horizon topic for the summit ? 21:44:43 just about right. I spend most of my Horizon time talking about cross-project stuff anyhow. 21:44:55 yeah 21:45:00 gabrielhurley: anything else ? 21:45:01 attendance was good, and the sessions were excellent. thanks for everyone involved! 21:45:27 can't think of anything urgent. I've got some stuff to bring tothe mailing list in the next week or so. 21:45:29 I really enjoyed the process ones I attended. For some reason there were not tha many people around and we had good discussions 21:45:46 +1 for being boring 21:45:51 Questions for Horizon ? 21:46:26 #topic Other Team reports 21:46:33 QA, CI: anyone ? 21:46:56 davidkranz: want to give some feedback from QA topic at the Summit, if around ? 21:47:19 ttx: Sure. 21:47:31 davidkranz: how did that go ? 21:47:43 Pretty well I think. The trick is the follow through. 21:48:01 Rackspace is going to upload all their tests soon. 21:48:08 davidkranz: did you get the people you needed on those sessions ? 21:48:16 and not too many of the people you didn't need ? 21:48:37 We didn't have many ptls, but we used one of Dan's sessions for the quantum stuff which was the most important. 21:48:51 davidkranz: we've already made progress on testtools/testr and I think you're going to like it 21:48:57 I think it's easier this way around 21:49:02 Dan Prince was interested in more use of SmokeStack. 21:49:11 mordred: That's great. 21:49:26 davidkranz: anything else ? 21:49:52 ttx: We also had some in-person QA meetings with summaries posted to the qa list. 21:50:00 ttx: That's about it. 21:50:08 Any other team lead with a status report ? 21:50:12 tons of sessions involved cI 21:50:14 https://etherpad.openstack.org/state-of-ci 21:50:20 there's what we collected overall 21:50:24 for those who are interested 21:50:32 also, if you didn't notice, we have a new etherpad server 21:50:38 mordred: I think the Process topic went well overall, what do you think ? 21:50:44 I agree too 21:50:59 mordred: people too fried to disagree with moderator 21:51:09 I do not really remember any contentious topics 21:51:34 mordred: I just rememberd that we need to discuss oslo versioning once markmc will be back 21:51:40 yes 21:51:51 i.e. aligned vs. tag vs. forward vs. pip 21:51:52 ttx: you set that as an action back in the oslo topic 21:51:58 yep 21:52:04 #topic Open discussion 21:52:12 #info We'll skip the meeting next week 21:52:22 As mordred knows, i'll have better things to do 21:52:35 and I don't expect the plans to be finalized anyway 21:52:46 So not so much to discuss in-meeting 21:52:50 see you all at UDS next week... 21:52:52 * mordred cries 21:52:56 I'll bug you all in the next two weeks so that we make progress on that though. 21:53:01 * russellb won't be there :-p 21:53:06 (that being, grizzly plans) 21:53:24 Anything else, anyone ? 21:53:42 Next meeting on Nov 6 21:54:04 WARNING WARNING we are entering DST confusion zone 21:54:12 Europe drops DST this weekend 21:54:33 Please triple-check meeting times (they are in UTC!) before going to one 21:54:57 #endmeeting