16:00:43 #startmeeting cinder 16:00:44 Meeting started Wed Mar 6 16:00:43 2013 UTC. The chair is jgriffith. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:45 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:47 The meeting name has been set to 'cinder' 16:01:14 Hmmm... anybody here this morning? 16:01:20 Hi 16:01:24 hi! 16:01:29 hi 16:01:30 y 16:01:42 There's some folks... morning guys 16:01:50 o/ 16:02:08 Pretty short agenda today 16:02:20 might be even shorter without some of the folks being around 16:02:28 morning! 16:02:34 hi 16:02:34 heyhey 16:02:43 alright, now we're rolling, let's get started 16:02:51 #topic RC1 status updates 16:03:14 I'm hoping to cut next week unless folks see a problem getting their fixes in 16:03:31 Or of course somebody finds something horrible between now and then 16:03:39 any objections on that? 16:04:00 BTW: https://launchpad.net/cinder/+milestone/grizzly-rc1 16:04:34 Take a look and if you have something new let me know, or if you have something that should be targetted and isn't let me know that as well 16:04:35 hey sorry I'm late 16:04:41 bswartz: no worries 16:04:42 i see john's got 6 bug assigned. :) 16:04:54 winston-d: :) 16:04:56 jgriffith: It is too late to get the LeftHand driver fixes into RC1 16:05:04 kmartin: nope 16:05:27 kmartin: I am not super strict about targetting 16:05:45 kmartin: but we should probably update those bugs with RC1 tags 16:05:57 ok, I 'll have hemna send bug to you today 16:06:04 kmartin: do you expect to get them in this week? 16:06:10 yes 16:06:17 kmartin: we should be just fine then 16:06:46 winston-d: that reminds me, do you want https://bugs.launchpad.net/cinder/+bug/1131322 16:06:48 Launchpad bug 1131322 in cinder "Cinder service not logging exceptions from stevedore library" [High,Confirmed] 16:07:04 jgriffith: yes, please 16:07:11 winston-d: All yours :) 16:07:13 jgriffith: Can we have this in rc1 as well? https://bugs.launchpad.net/cinder/+bug/1131291 16:07:15 Launchpad bug 1131291 in cinder "XenAPINFS: Volume always uploaded as vhd/ovf" [Undecided,Fix committed] 16:07:15 jgriffith: https://bugs.launchpad.net/cinder/+bug/1136174 has been fixed aswell so not sure if you want to target that too? 16:07:17 Launchpad bug 1136174 in cinder "cinder-backup doesn't use/check metadata version during restore" [Undecided,Fix committed] 16:07:21 winston-d: do you think we can get that done this week? 16:07:27 jgriffith: but if it requires something in olso, it might not make it to RC1 16:07:59 matelakat: committed ones don't need much persuasion frm me :) 16:08:20 smulcahy: I can if you'd like it for tracking/record keeping 16:08:35 smulcahy: done 16:08:39 jgriffith: i'll try my best 16:08:44 jgriffith: ta 16:08:45 winston-d: I was afraid of that 16:08:57 jgriffith: is there an update from j_king on https://bugs.launchpad.net/cinder/+bug/1087817 ? 16:08:58 Launchpad bug 1087817 in cinder "Update v2 with volume_type_id switch to uuid" [Medium,In progress] 16:09:00 winston-d: Let me know if I need to work with Mark to try and get it in if you can get it proposed 16:09:08 jgriffith: So I don't need to take actions to get that to rc1, right (sorry, if this is a stupid q) 16:09:15 jgriffith: sure 16:09:43 thingee: I've not heard back from him, I did ping him the other day on it 16:10:08 I can do last minute work if needed there 16:10:10 thingee: wonder if we should just reassign and fix it 16:10:12 :) 16:10:16 I'm sure he has it though 16:10:21 thingee: alright, let's give him another day or so 16:10:26 thingee: thanks! 16:10:36 matelakat: nope, I already did it 16:10:50 and just in case others are not sure how that works... 16:11:01 basicly anything that's going in to trunk right now is going in to RC1 16:11:28 Oh, ok. 16:11:35 The targetting exercise just makes sure we know what we have deemed as *required* for rc1 16:12:25 Technically one could argue that everything that goes in should be targetted at this point 16:12:38 but Cinder is a bit unique with it's heavy driver model :) 16:13:02 Any folks having the chance to do some good testing lately? 16:13:43 I'd like to see as many of us as possible start switching context to testing and documentation over the next few days if possible 16:14:11 and grenade testing as well (Folsom to Grizzly upgrade testing) 16:14:42 Well I see that was a popular topic :) 16:14:44 btw, I asked Anne About it, but is this the right place for the driver docs? #link http://docs.openstack.org/trunk/openstack-compute/admin/content/ch_volumes.html 16:15:17 matelakat: it is for now, I have a todo item to create a "block storage" section and move all of that out of compute admin 16:15:47 #action jgriffith get the doc formatting stuff figured out with Anne and done 16:16:19 any other big questions/concerns for RC1? 16:16:51 #topic core team status 16:17:11 So we've talked about this before but always wait for the *next* mile marker 16:17:30 I'd like to propose that we clean out the core list of cinder 16:17:52 What I mean here is the folks that were set up in Folsom but never really participated 16:18:06 jgriffith: is this "core list" published somewhere? 16:18:14 I'd like to trim that list and make some new nominations this week if possible 16:19:11 bswartz: yes of course https://launchpad.net/~cinder-core 16:19:34 thanks 16:19:40 what do you all think about starting a storage admin manual? 16:19:51 annegentle: :) 16:20:06 also I would like eyes on https://review.openstack.org/#/c/23625/ 16:20:08 Do we have a list of contributers and reviewers for the last few months? Seems like a good place to start for list trimming... 16:20:12 annegentle: That's what I was getting at earlier 16:20:27 sorry I am late and the topic changed :) 16:20:33 annegentle: no worries 16:20:45 annegentle: I'll check it out :) 16:20:50 I'll have a look at the review you pointed out as well 16:21:07 annegentle: and I'd definitely like to have a block-storage admin guide 16:21:15 move the cinder stuff out of the compute admin doc 16:21:22 jgriffith: yeah I think that's the way to go 16:21:32 jgriffith: anyone in mind to do the refactor? 16:21:47 annegentle: unless somebody else steps up proabably me 16:22:07 jgriffith: there may be others interested I can recruit for you if you want :) 16:22:08 annegentle, jgriffith: I did it for api, why not? :) 16:22:11 I'd like this to be ASAP 16:22:15 thingee: woo! :) 16:22:25 thingee: yeah you have your plate full getting the API docs in shipshape 16:22:29 thingee: You're the MAN!!! 16:22:56 So let's do this 16:22:59 those are at least in a review phase. I still have more I want to do with them, but in the mean time can focus on admin 16:23:14 I'd like to have this refactor by next week at latest 16:23:35 If anybody has some cycles and gets to it before I'll have a chance (which will be mid next week) 16:23:45 Let folks know 16:24:00 We'll open a docs bug for it and target it so folks will know if it's picked up active etc 16:24:08 annegentle: thingee sound good? 16:24:12 jgriffith: sounds jus tright 16:24:15 just right even 16:24:18 coolness 16:24:23 sounds good 16:24:34 jgriffith: is there any officially organized group of "cinder team members" who are not not in the core team but more involved than your average Joe on the street? 16:24:37 thingee: if you've got the band-width that's awesome by me :) 16:24:40 I don't have anything planned atm, so I'm fine with it 16:24:45 bswartz: nope 16:24:59 bswartz: DuncanT to that point searches in gerrit are the answer 16:25:28 jgriffith: i'll see if i can provide you some review numbers tomorrow 16:25:32 jgriffith: I'll try to produce some stats if you want... 16:25:45 (or winston-d can beat me to it... I'm easy ;-) ) 16:26:01 DuncanT: :) 16:26:07 winston-d: oh, cool I was going to be less formal and do it ad-hoc 16:26:20 winston-d: however the stats would be good for *new* core nominees 16:26:32 jgriffith: is there are difference between the "cinder core" team and the set of people who are "core approvers" (can +2 stuff)? 16:26:46 bswartz: nope, they're one in the same 16:26:57 okay 16:27:34 anybody have anything else on core team, or docs? 16:27:34 jgriffith: Regarding docs, will they take changes all the way up to the end of the release or are the docs connected to the RC releases as well 16:27:52 kmartin: docs we have more flexibility there 16:28:16 but we've procrastinated before and we always seem to run out of time 16:28:25 kmartin: right. We take a look at the bugs and figure out if the docs are releasable 16:28:38 ok we have doc person that wants to be a OpenStack contributor, so he is working though that process 16:28:39 kmartin: it's that balancing act between "will this answer questions or cause more questions" 16:28:48 kmartin: NICE! 16:29:51 alright, folks think about if you know of somebody that you think should be core, or if you want to nominate somebody (even yourself) and let me know 16:29:55 goal is to get the changes in the end on next week 16:30:01 regarding the core team -- I find the idea of a 2-tier core team structure appealing 16:30:38 bswartz: I'm not really interested in changing the OpenStack management/organization structure but I'm listening... 16:31:10 bswartz: what are you proposing? 16:31:37 well I sympathize with the desire to keep the team small, and weed out non-participators, but I think it's good to have a wider group of people who are incentivised to be online answering questions and participating in code reviews, etc 16:31:58 bswartz: Actually that's what *core* is supposed to be 16:32:28 bswartz: our team is rather small because our consistent participation until the past couple of months has been relatively minimal 16:33:02 I guess I worry that people who don't make the cut for the "core team" won't feel any incentive to contribute 16:33:05 bswartz: and everyone so here's my metric... 16:33:15 What I've been looking at the past week is this: 16:33:29 1. People that have contributed code to the core cinder project this cycle 16:33:39 2. People that have reviewed code 16:33:55 3. People that participate in IRC (including this meeting) 16:34:56 maybe your definition of core team includes my notion of both tiers or participation 16:35:04 bswartz: perhaps 16:35:16 Also bear in mind core definition is an OpenStack wide policy 16:35:18 I just wanted to throw it out there -- I haven't though this through fully 16:35:18 https://wiki.openstack.org/wiki/Governance/Approved/CoreDevProcess 16:35:30 thought* 16:36:01 bswartz: believe me, I'm all for having more core team members! 16:36:19 and I plan to nominate at least 3 folks later today 16:36:37 bswartz: and if you'd like to step up and take on the responsibilities let me know 16:36:56 as an occasional contributor to various free software, I would be very motivated to be on a separate list. I understand the separation between "core" and "contributor" 16:37:07 my 2ct ;-) 16:37:17 dachary: I'm confused, we already have this 16:37:26 dachary: bswartz So we have contributors 16:37:38 contributors can review, provide input, do anything they want 16:37:44 jgriffith: yes, I was echoing on the "no motivation if not in the core" ;-) 16:37:54 core just means you're especially active and responsible for +2/A authority 16:38:09 dachary: Ahh... ok, :) 16:38:11 thanks 16:38:46 My thought is if you're not motivated when you're not core, you won't be much of a core contributor anyway 16:38:53 but I digress 16:39:16 bswartz: If you have some structuring ideas we can definitely talk through them and see if there's a better system 16:39:55 anything else regarding core? 16:40:12 #topic summit session 16:40:44 http://summit.openstack.org 16:40:54 jgriffith: when is deadline for submitting session for summit? 16:40:57 #link http://summit.openstack.org/cfp/details/38 16:41:23 winston-d: I actually don't have an answer on that for you 16:41:37 We haven't even begun talking about track scheduling etc yet 16:41:58 jgriffith: k. i'll submit my ideas ASAP anyway 16:42:22 I've discussed this tentative session on the ceph mailing list and there is an interest. I feel it's better associated with cinder but I'm not 100% sure. Roadmap for Ceph integration with OpenStack http://summit.openstack.org/cfp/details/38 16:42:25 winston-d: good idea :) 16:42:32 If nothing else submit things as place holders 16:42:49 dachary: yeah, I've been following that 16:43:23 jgriffith: what's your advice ? 16:43:36 dachary: I have no good advice at this point :) 16:43:41 dachary: So my take was/is 16:44:01 if there's room in Cinder tracks I don't mind it, however what you have planned seems much more nova related IMO 16:44:26 dachary: bottom line, if it needs a place to land and you're in a jam we can work it out in Cinder 16:44:51 but it almost seems you need your *own* independent track 16:45:09 jgriffith: +1 16:45:11 dachary: or if we do another *general all/around* bucket again that would be ideal 16:45:11 that makes sense, I'll move it to nova then. 16:45:42 dachary: Your call, if you get in a bind scheduling wise let me know and we'll work something out 16:45:53 jgriffith: thanks :-) 16:45:54 dachary: I also think we should ask about cross-project tracks 16:46:32 I'll discuss this with ttx, good idea 16:46:40 cool 16:47:03 we've had that sort of thing in the past IIRC 16:47:15 Ok... anything else on sessions? 16:47:17 bswartz: ? 16:47:38 I'm assuming we're going to need to hash something out for real this time on File Shares 16:48:29 bswartz: But that might be better as it's own type of session as well (non-cinder track) 16:48:40 well if nobody else has anything.... 16:48:43 We're planning to address the feedback we've gotten so far on the NAS stuff so it's ready when Havana opens up 16:48:48 jgriffith: Would you rather have separate proposals for small things then combine them later? 16:49:00 DuncanT: small and combine later please 16:49:08 dachary: IMO, ceph has little to do with cinder's core. Unless your proposed changes are actually requesting big changes. 16:49:12 DuncanT: it tends to make things easier to schedule 16:49:19 thingee: +1 16:49:20 jgriffith: Understood 16:49:40 DuncanT: but if you have a well organized grouping by all means, it's up to you 16:49:56 I worry a little that another conference session might feel like a rerun given that the features have not changed significantly, just the impelmentation 16:50:04 DuncanT: It's just easier for me to see overlap and tie ins if they're seperated 16:50:10 thingee: yes, you are correct 16:50:20 bswartz: agreed, I have some suggestions for how to do this differently this time 16:50:39 bswartz: one is already taken care of (don't have a *surprise* proposal in the session) 16:50:58 bswartz: I think some of the concerns, particularly about buy-in from other users, might benefit from a session if we can get some of them there 16:51:05 if there had been some communication and the code was shared before the session I think it would have been moe productive 16:51:23 s/moe/more/ 16:51:44 jgriffith, DuncanT: I agree 16:51:57 DuncanT: bswartz Recall in SanFran there was actually a storage panel that discussed this with a broader audience 16:52:13 The resounding response was *yes, we want shared filesystem support* 16:52:20 *no, it shouldn't be in Cinder* 16:52:34 It's actually on video somewhere 16:52:39 jgriffith: I'm not sure which session you're referring to 16:53:07 bswartz: It was a panel discussion... Rob, SF, Mirantis, Nexenta 16:53:29 On the first day? I think I was there 16:53:42 bswartz: I think it was the last day, but not sure 16:53:46 I have a different memory of events -- but I might be remembering a different session 16:53:48 I'll see if I can't find the link 16:53:52 please do 16:54:23 Yes you and Rob have pointed out that my memory is apparantly *wrong* 16:54:46 I think the point of contention is about how resounding the response was 16:55:02 clearly some people think NAS doesn't belong in cinder 16:55:36 bswartz: sure... 16:55:45 anyway, we'll sort it this time around 16:55:55 anybody have anything else? 16:56:08 I'll find the link and hit you up in the channel with it 16:56:17 Ok... thanks everyone 16:56:24 The work regarding multi backend / tempest integration test is in progress. 16:56:25 #end meeting 16:56:25 thanks :-) 16:56:31 #endmeeting