21:03:26 #startmeeting 21:03:27 Meeting started Tue Sep 6 21:03:26 2011 UTC. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:03:28 Useful Commands: #action #agreed #help #info #idea #link #topic. 21:03:42 Welcome to our weekly team meeting... Today's agenda is at: 21:03:45 #link http://wiki.openstack.org/Meetings/TeamMeeting 21:03:52 quite a lot to go through today 21:03:58 #topic Actions from previous meeting 21:04:05 * vishy to email the list about the other critical bug and request help on reviews and bug fixing 21:04:19 did that 21:04:33 #topic Swift status 21:04:41 swift moved to gerrit today 21:04:50 notmyname: you confirm we'll delay 1.4.3 for a couple days ? 21:04:56 and the 1.4.3 release is pushed back until friday/monday 21:05:37 Looking at: https://launchpad.net/swift/+milestone/1.4.3 -- looks pretty ready to me 21:05:49 notmyname: anything else ? 21:05:53 nope 21:06:01 Raise your hand if you have questions on Swift... 21:07:45 #topic Glance status 21:07:49 jaypipes: o/ 21:07:49 https://launchpad.net/glance/+milestone/diablo-rbp 21:07:59 On the features side, looks like you still need to land 3 blueprints ? 21:08:03 lots of bug fixing to do. most features OK. 21:08:11 * support-ssl (test fails) 21:08:15 * pluggable-auth (needs fixing) 21:08:21 * protected-properties (blocked by pluggable-auth) 21:08:22 ? 21:08:28 the user- and system-controlled properties should now be unblocked. need to get with Vek to discuss who should handle that... 21:08:59 ttx: or at least, unblocked when pluggable-auth is fixe.d 21:09:22 ttx: there is some talk of API extensions needed, but I'm pretty sure that will not be done in Diablo. 21:09:43 ttx: and will be done early in Essex, with the main split of the API endpoint into a registry and mover API... 21:09:43 jaypipes: those features need to land now... any ETA ? 21:09:44 jaypipes: I'm committed this sprint, so I don't think I'll be getting to it... 21:10:21 ttx: give me a day to talk with pvo and westmaas about bandwidth for the user- and system-properties one. other ones should be OK. 21:10:25 ttx: for the features. 21:10:37 jaypipes, we have the system proprerties scheduled for our sprint that starts next week 21:10:50 glenc: that's a bit late for me 21:10:54 glenc: might be too late. 21:11:13 glenc: we would need to get started on it earlier than that 21:11:15 Last stop for features is RBP 21:11:28 you're already on borrowed time 21:11:34 glenc: which means done by 8th... :( 21:12:19 ttx: I will get with glenc offline on that. 21:12:23 ok 21:12:23 please 21:12:32 jaypipes: On the bugs side, you have 10 targeted bugs 21:12:42 I guess I'll just push them to "2011.3" if they don't get fixed in time for RBP ? 21:13:10 ttx: the critical and High I really want in. Medium to Low may get pushed. 21:13:18 ok, noted 21:13:22 Note that starting at RBP only bugs that are targeted to release will get backported to the release branch 21:13:33 So don't hesitate to ping the PTL or me to make sure essential bugs are covered. 21:13:41 jaypipes: Other announcements/comments ? 21:14:23 ttx: yes, understood. 21:14:30 Raise your hand if you have a question on Glance. 21:15:00 ttx: yes, will do. thx. last week was a bit of a complete loss for jeblair and myself :( 21:15:16 jaypipes: yes, I saw that :/ 21:15:32 bad timing 21:15:41 #topic Nova status 21:15:53 vishy: hey 21:15:58 yo 21:15:59 Looking at: https://launchpad.net/nova/+milestone/diablo-rbp 21:16:05 3 "features" still on the roadmap: 21:16:12 * implement-network-api (needs review) 21:16:19 two approves 21:16:27 Apparently all comments have been addressed -- _cerberus_, _0x44: could you doublecheck ? 21:16:29 dan should be good 21:16:36 I believe the feedback for the other two reviews have been incorporated, but waiting on confirmation from the reviewers 21:16:41 * aws-api-validation (needs fixing) 21:16:44 * hyper-v-update (in progress) 21:16:53 vishy: I propose to convert those last two to bugs, as they are not really "features" -- would that work for you ? 21:16:54 hyperv we push (havent' heard anything from jordan) 21:17:03 <_0x44> ttx: I just responded again to aws-api-validation 21:17:08 validation is a bug for sure 21:17:39 hyperv's description is a bit scary in corners 21:18:19 ...changes to nova-compute and service.py... 21:18:30 but without code linked it's a bit hard to know 21:18:58 if it really makes changes to service.py I'd like it to land before RBP (or never) 21:19:08 yeah i don't think it is coming in 21:19:18 so i think essex is probably fine 21:19:27 <_cerberus_> ttx: yeah, i can check again 21:19:54 vishy: will refurbish as bug 21:19:59 k thx 21:20:33 #action ttx to refurbish aws-api-validation and hyper-v-update as bugs and untarget them 21:20:42 stool don't have a person to do: https://bugs.launchpad.net/bugs/834189 21:20:42 Launchpad bug 834189 in nova "XenServer and KVM handle local storage differently" [Critical,Triaged] 21:20:54 yes, you've 6 targeted bugs, and 4 of them are unassigned 21:21:00 <_0x44> ttx: I'll look at it also 21:21:03 I think allowing an option for kvm to do the resize like xen is good enough 21:21:28 vishy: lower Importance, maybe ? 21:21:51 i guess we could go down to high. I really think that should be in for release though 21:21:58 We also need assignees for: 21:22:07 https://bugs.launchpad.net/bugs/833331 21:22:08 Launchpad bug 833331 in nova "OSAPI v1.1 needs to document config-drive as an extension" [Medium,Confirmed] 21:22:19 https://bugs.launchpad.net/bugs/827807 21:22:21 Launchpad bug 827807 in nova "in multi_host nova-network mode, nova-network doesn't reassociate reassociate the floating ips on reboot" [Medium,Triaged] 21:22:25 https://bugs.launchpad.net/bugs/838386 21:22:27 Launchpad bug 838386 in nova "Test suite requires kombu and carrot to be installed" [Low,Triaged] 21:22:32 Any takers ? 21:22:57 <_0x44> ttx: 833331 should be me, probably. When is RBP due by? 21:22:57 I could take that last one. 21:23:08 i thought cbehrens had a fix in for that last one 21:23:19 ttx: i think Daviey was working on that somewhat (838386) 21:23:23 _0x44: Thursday. that can be fixed after RBP though 21:23:26 he was definitely discussing it with Daviey in the channel 21:23:35 i gave him a fix.. 21:23:35 ah there you are forgot u were comstud here 21:23:37 _0x44: will just require a backport to the release branch. 21:23:41 doesn't appear he's linked a branch or claimed the bug 21:23:52 vishy :) 21:23:56 _0x44, soren: please assign yourselves 21:24:08 ttx: Will do and will talk to Daviey tomorrow. 21:24:19 vishy: anything else ? 21:24:25 nope 21:24:29 Questions for Nova PTL ? 21:24:37 * Vek has known comstud as comstud for a decade or two... 21:24:40 Hi PLZ review https://code.launchpad.net/~ntt-pf-lab/nova/single-default-gateway/+merge/74159 21:24:54 We fixed some bugs 21:25:00 vishy: do we really want that in Diablo ^ ? 21:25:15 it is in RDP milestone 21:25:34 It was added last meeting 21:25:42 oh 21:25:50 I see :) 21:26:06 And also it is important bug fix when we use multi nic 21:26:07 then yes: ^ review early please 21:26:21 #topic Incubated projects news 21:26:22 so please review it :D 21:26:44 devcamcar, danwent: o/ 21:26:52 quantum is focusing on rbp…. adding more unit tests 21:26:55 anyone from keystone around ? 21:26:57 ttx: dashboard isn't incubated anymore :) 21:27:04 devcamcar: in fact, it is. 21:27:04 no major features are outstanding 21:27:12 devcamcar: we are still in Diablo :P 21:27:17 ttx: touche! 21:27:43 we are building a presentation for essex summit to outline what we'll be up to 21:27:44 quantum: keystone integration is also being added during rbp 21:28:01 devcamcar: had a question for you btw -- does Dashboard support more than Nova and Quantum at this point ? 21:28:15 if people are planning network related blueprints for essex summit, please keep us in the loop 21:28:22 we'll be focusing on our approach to user interaction and the revamps both visually and in flows for dashboard 21:28:26 devcamcar: how is work going on the doc updates noted in https://bugs.launchpad.net/openstack-manuals/+bug/813072? 21:28:27 Launchpad bug 813072 in openstack-dashboard "Dashboard documentation needs to be updated for changes in configuration relating to Keystone integration." [Critical,In progress] 21:28:50 ttx: we have support for swift as well (and glance implicitly) 21:29:02 we'll be adding some features that are glance specific soon, for uploading images, etc. 21:29:08 devcamcar: great! 21:29:13 annegentle: starting next week we're going to throw some resources at that 21:29:30 devcamcar: great, throw them with good aim :) 21:29:40 devcamcar, danwent: anything else ? 21:29:45 ttx: nope 21:29:55 we'll be working on docs too…. let me know if you want to help :) 21:29:56 nothing else 21:29:58 #topic Proposing sessions for the Design Summit 21:30:13 First, if you haven't registered for the summit yet, the waiting list is still open but will close sometime this week 21:30:20 or next week 21:30:26 So please make sure you register at http://summit.openstack.org -- we try to prioritize known contributors 21:30:39 For registered attendees, now it's time to propose sessions ! 21:30:47 For the main tracks, we'll have 3 different session types: 21:30:58 * Brainstorm sessions (55 min.) - used to discuss and come up with a solution for complex issues 21:31:10 * Rubberstamp sessions (25 min.) - used to present and review an already-designed plan 21:31:22 * Discovery sessions (25 min.) - where experts go into deep detail into a section of code or feature 21:31:38 that last type is new ^ 21:31:39 /me wonders which is the 'brawling' session 21:31:43 :) 21:31:50 Anyone can go to http://summit.openstack.org to propose session topics (click on "Proposed Sessions") 21:32:05 Note that we no longer *require* the creation of a blueprint for every session 21:32:20 This is only useful if the session talks about a specific feature / code drop and you already have a plan 21:32:36 So blueprints should be filed in support of "Rubberstamp" sessions (you can link to them from the session proposal) 21:32:48 ttx: sorry, lost internet for a minute 21:32:55 Sessions will be reviewed by PTLs for each track and myself, check back the site to see status and comments. 21:33:13 vishy: you didn't miss that much 21:33:23 Details at: http://wiki.openstack.org/Summit 21:33:29 Questions ? 21:33:59 ttx: in regards to that branch, I would love comments from pvo / tr3buchet as they are using multinic quite a bit. I haven't needed it yet 21:34:26 ttx: deadline for submission and is there a deadline for approval? 21:34:27 I don't see a problem pushing it to essex unless it is really causing issues for people that will be stuck on diablo for a long time 21:34:40 blamar: let me check the timeline 21:34:49 ttx: although it could be considered a bug fix, although it is perhaps a bit risky 21:35:38 Propose sessions (Sep 6 - Sep 27), Review sessions (Sep 12 - Sep 29), Schedule sessions (Sep 27 - Oct 2) 21:35:51 ttx: ty 21:35:52 blamar: so at least until Sep 27... 21:36:04 blamar: but to increase your chances you should submit early 21:36:10 vishy: sorry,which branch? 21:36:37 no other questions ? 21:37:05 #topic Discuss creating an OpenStack Common package 21:37:10 blamar: floor is yours. 21:37:47 Basically just wanted to talk about OpenStack common, and what needed to be done to get a package made for it in PyPi and Ubuntu 21:38:05 Not something I just wanted to "do" without at least bringing it up, and seeing if it was something that needed to go through the PPB 21:38:24 pvo: the single_default gateway above 21:38:26 The level of code duplicate across Glance, Nova, and Keystone is pretty bad. 21:38:26 * Vek thinks we need something like this 21:38:50 the melange merge is already starting to use it 21:38:57 yeah, lot of DRY violations... 21:39:07 or at least, what's available at the moment in openstack-skel 21:39:22 jk0: That's true, but the other part of it is that I *don't* want openstack common to just be copy/paste from Nova 21:39:31 jk0: it should be tested, and re-thought-out 21:39:32 right, I'm just saying 21:39:48 melnage would be a good place to test it 21:39:56 they are willing/already trying to do this 21:40:02 jk0: I'd really like it to be difficult to get code into OpenStack Common..because it would need to be tested and be the pristine copy 21:40:45 makes sense... 21:41:23 I'm very much in favour of an openstack-common sort of thing as well. 21:41:46 Anyway, I'm not sure really where to go from here. I have a branch which is bare-bones based off Jay Pipes common 21:41:48 https://code.launchpad.net/~blamar/openstack-common/my-common 21:42:24 It's basically nothing but some SQLAlchemy common code, but I'd like to start working with openstack-common as soon as Essex opens 21:43:06 So maybe I can throw some emails/review requests out to the list and gather feedback on guidelines for including stuff in Common? 21:43:38 blamar: sure -- it's not the first time that was pushed though -- the issue has always been resources to commit to it 21:43:41 It would be good to gauge PTL buy-in to this. 21:44:05 No point in building it if noone wants to change their respective projects to use it. 21:44:16 #action blamar to discuss openstack-common on the ML 21:44:19 ttx: Absolutely, at this point I truly believe we're wasting more resources on code duplication than we would on diverting effort to openstack-common 21:44:21 (which IIRC is why it didn't go anywhere last time somone tried to push it) 21:44:32 * glenc thinks someone should take the PTLs out back and "explain" it to them 21:44:55 * Vek knows a couple of black belts... 21:44:59 blamar: anything else ? 21:45:01 I don't think we had PTLs the last time this was seriously considered 21:45:09 ttx: done, thanks! 21:45:15 * vishy is bought-in 21:45:28 #topic Discuss creating API doc repos 21:45:33 annegentle: shoot! 21:45:33 jk0: Not formally, perhaps. 21:45:38 bang! 21:45:42 aw 21:45:47 just wanted to circle back here on my proposal to the mailing list today 21:45:48 Texans. 21:45:53 :) 21:46:05 I think I'm understanding the review process better, thanks to notmyname et al 21:46:06 hee 21:46:32 it appears that we could have a openstack-doc-core team for reviews, which I think is nice 21:47:01 and, separating out the API docs makes the main manuals project much less "build heavy" 21:47:19 so I hope y'all (that's Texan, right?) are okay with the move to Github and separation. 21:47:26 annegentle: I'm not sure on the delineation between using the project code name and the ...other...name? Why name it openstack/nova-api and not openstack/compute-api? 21:47:45 blamar: ah, good point! Especially for the api, compute-api makes perfect sense 21:48:00 identity-api? storage-api? 21:48:02 annegentle: I feel like we use the codename too much, but I might be alone on that. 21:48:12 auth-api, image-api, storage-api? 21:48:15 blamar: no, you're onto something 21:48:30 and image-api 21:48:54 #action annegentle to email the mailing list with new naming possibilities 21:49:29 annegentle: anything else ? 21:49:31 anything else? 21:49:34 ha beat me to it :) 21:49:49 just wanted to make sure I've covered the bases. Or some other Texas metaphor. 21:50:12 that dog'll hunt 21:50:26 haha :) 21:50:29 LOL 21:50:45 #topic Open discussion 21:51:12 So I heard temperature finally dropped in Texas ? 21:51:25 sure, it's 91 today. 21:51:29 I had a goosebump this morning 21:51:37 yes! 21:52:11 anything, anyone ? 21:52:42 hope everyone had a good labor day/equivalent holiday, if they had one... 21:54:04 ok then. That's all for today 21:54:13 #endmeeting