21:01:27 #startmeeting nova 21:01:28 Meeting started Thu Nov 14 21:01:27 2013 UTC and is due to finish in 60 minutes. The chair is russellb. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:01:30 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:01:32 The meeting name has been set to 'nova' 21:01:32 hello, everyone! who's around? 21:01:38 o/ 21:01:41 o/ 21:01:41 hi 21:01:41 hi 21:01:42 hi 21:01:43 o/ 21:01:43 o/ 21:01:44 hi 21:01:44 hi 21:01:46 hi 21:01:48 hi 21:01:54 hullo 21:01:55 awesome, let's get going then 21:02:04 #topic general announcements 21:02:10 #link https://wiki.openstack.org/wiki/Meetings/Nova 21:02:12 that's our meeting page 21:02:16 i'll be keeping an agenda on there each week 21:02:22 if you have something to add, feel free to edit the page, or ping me 21:02:23 o/ 21:02:30 also, if you have ideas for things that we should hit every week, let me know 21:02:45 Hi 21:02:52 the other general thing for today is making sure everyone has seen the icehouse schedule 21:02:58 #link https://wiki.openstack.org/wiki/Icehouse_Release_Schedule 21:03:07 #note icehouse-1 release is 3 weeks from today 21:03:16 the release will fly by :) 21:03:35 so, 66 blueprints targeted at icehouse-1, and 0 to the other milestones isn't terribly realistic, but that's another topic :) 21:03:37 holy crap 21:03:41 dansmith: exactly 21:04:21 so as you're planning your blueprints, please look closely at the schedule and figure out what's realistic for your work 21:04:26 i'll go ahead and say instance type flavor dedup is probably I2 21:04:26 so we have to merge 33 BPs a week 21:04:28 russellb: how firm is that May 15th week for the next summit? 21:04:38 err 22 21:04:40 mikal: contract signed AFAIK 21:04:48 mikal: so very firm 21:04:52 russellb: cool 21:04:58 russellb: its nice to be able to plan 21:05:14 so, week of May 12-16 21:05:22 i think they have 5 days booked 21:05:26 and we may offset design summit by a day 21:05:34 so something like, conf mon-thurs, design summit tues-fri 21:05:38 or something, not positive yet 21:05:39 Cool 21:05:40 but very likely 21:05:55 Do you have dates for the mid-cycle meetup? It seems like that might be the week of icehouse-2? 21:05:56 "they" being the foundation, that organizes these things 21:06:15 ok, so mid-cycle meetup, good question 21:06:34 i gave some possible dates to the sponsors (bluehost), and last week they told me to just pick and week and they'll make it work 21:06:38 so now it's back on me 21:06:39 or us 21:06:44 Pick me! 21:06:47 I can be the decider 21:07:08 heh 21:07:14 ok, so based on the poll i did a while back 21:07:21 good weeks looked like either january 13th 21:07:34 or feb 17 21:07:50 * dansmith votes for the latter 21:07:50 Jan 13 will clash with the ATC meetup in Australia 21:07:53 feb 17 is a bit more "mid cycle" 21:08:13 I'm in favor of Feb as well 21:08:15 location? 21:08:20 in the US somewhere 21:08:21 Are you thinking an entire week? 21:08:24 depends on the date we tell them 21:08:29 Or like three days? 21:08:33 mikal: sounds like they're willing to host more like 3 days max 21:08:36 2 or 3 days 21:08:45 * beagles votes for St. John's 21:08:46 provide us space, and lunches 21:08:50 I think that's more realistic anyway 21:08:53 in terms of travel 21:08:57 Oh, so somethign like Feb 17 to 19? 21:08:58 Atlanta 21:09:05 geekinutah: around by chance? 21:09:05 sorry, 21:09:08 yep 21:09:16 possible locations sounded like Boston, or Phoenix 21:09:24 * russellb doesn't really care 21:09:25 yeah Boston Phoenix or Provo 21:09:32 * mikal wanted Utah 21:09:32 phoenix, warm 21:09:40 Provo means possible ski trip :) 21:09:42 I vote provo, phoenix, boston, in that order 21:09:46 this is true :-) 21:09:56 Where is Provo? 21:09:59 i haven't been snowboarding in years 21:09:59 Atlanta in Jan/Feb == coooooooold 21:10:01 mikal: utah 21:10:17 but any would be good :) 21:10:23 wondering if staying away from the east coast in that time of year would be a good idea ... for that kind of thing anyways 21:10:39 OK, so week of feb 17? 21:10:39 Oh yeah, airport closures 21:10:41 snowstorms messing with travel, etc 21:10:42 Atlanta was typo, sorry 21:10:43 yay to boston :) 21:11:02 russellb: my chances of getting you to run it a week earlier are zero, right? 21:11:04 Provo or Pheonix would be my suggestion 21:11:17 mikal: a week earlier hits valentines day, which some people care about 21:11:22 russellb: I have a probably clash depending on how long the meetup is for 21:11:41 What are the dates for Valentine's day? 21:11:42 unless we did a monday-tuesday meetup, perhaps 21:11:44 feb 14 21:11:46 (Don't tell my wife I asked that) 21:11:53 * russellb doesn't care about it, but some do 21:11:53 :D 21:11:56 +1 for avoiding valentine's day 21:11:58 comstud told me he'd get shot 21:12:03 mikal: that comment goes on facebook ;-) 21:12:03 for example :) 21:12:05 yeah and we need comstud 21:12:06 Well, 10 - 12 Feb would let people get back in time for the 14th 21:12:13 changing +1 to -1 to avoid getting comstud shot 21:12:14 And then I would miss my kids birthday... 21:12:27 mikal: what would make you miss it? 21:12:32 s/would/wouldn't/ 21:12:37 oh, i see 21:12:47 anyone *against* feb 10? 21:12:49 I need to be back in AU by 21 Feb, which makes the week of the 17th tight 21:13:05 mikal: 17-18 meetup? 21:13:10 would get you back ... barely 21:13:21 17 - 19 would get me back on the 21st 21:13:25 dansmith: i know you had some conflicts, feb 10 work for ou? 21:13:33 So tight but doable 21:13:37 * russellb nods 21:13:39 russellb: it would be tight for me, I'm looking 21:13:42 k 21:14:09 yeah, I could do it 21:14:15 ok, well, right now it's either feb 10 or feb 17, leaning toward feb 10, but i'll wait for anything further last minute "please no" comments, and i will re-review the poll results 21:14:25 from there i'll confirm it with bluehost, and report back 21:14:26 Cool 21:14:29 mikal: any chance you won't get funding, or will you definitely be there? 21:14:30 i know this needs to get firmed up asap for planning 21:14:34 sorry i'm terrible at this 21:14:38 I personally think three days would be better than two by the way 21:14:50 mikal: OK 21:14:52 dansmith: I need to check... I can do that today though. 21:15:02 Do we have a mail thread for this I can let people know on? 21:15:04 mikal: i know 2 is quick for how far you're traveling ... 21:15:13 mikal: okay, if it's unlikely, then maybe we could push for the later, and if not, then the earlier 21:15:24 but maybe a smaller group can hang out another day, i dunno 21:15:24 i don't want to overextend our welcome 21:15:28 I think the mean consensus for length was 3 days also 21:15:32 mikal: no thread right now, please ping me 21:15:39 russellb: sure 21:15:56 geekinutah: yeah, but jsmith acted like 3 might be a bit much 21:16:04 i expect my employeer to sign the stupid CCLA. if they do i hope to be more involved and I will attend. I would love a snowboarding the weekend after with whomever is up for it. 21:16:06 we should check with him on that 21:16:13 I think it can be done 21:16:19 jaybuff: awesome 21:16:22 we just need to move schedules around 21:16:25 geekinutah: OK 21:16:39 snowboarding fun 21:16:41 we are 15 min into the meeting and 3 weeks away from I-1 ... 21:16:42 geekinutah: do we get a tour of facilities? 21:16:49 certainly 21:16:51 so ... to firm this up, i'm leaning toward feb 10 based on mikals conflict 21:16:53 sweet 21:16:56 and will ask for 3 days, but we'll happily take 2 21:17:04 +1 provo, feb 10, 3 days 21:17:07 going to wait a day for mikal to get back to me 21:17:13 and then will push forward 21:17:22 sound good? 21:17:42 Thanks guys 21:17:45 * russellb nods 21:17:49 ok, onward 21:17:50 #topic bugs 21:18:02 one thing i've mentioned by mail, and in the design summit, is that we need a bug czar 21:18:15 because it's one thing that i drop the ball on trying to do everything 21:18:22 lifeless spoke up at the design summit as willing to start helping with this 21:18:43 for reference, here is our current triage process: 21:18:48 #link https://wiki.openstack.org/wiki/Nova/BugTriage 21:18:52 however, 82 bugs are untagged 21:19:01 and 100 tagged bugs are untriaged (181 total untriaged) 21:19:05 so, we aren't doing great at it 21:19:27 one person owning triage gets hard, i've done that before on projects and we had to start rotating it 21:19:29 lifeless: around? thoughts for how we can improve on our bug queue handling? 21:19:33 hi 21:19:38 nuts, today is a public holiday 21:19:40 well i envision a person leading the effort, building a team 21:19:52 the bug thing is at the top of my todo list but I vagued on it cause I wasn't going to be here. 21:19:58 I do have a plan! 21:20:07 lifeless: ok, all good, shall we revisit next week? 21:20:10 three second over view 21:20:17 i like plans 21:20:21 k! 21:20:25 get some metrics that I think are valuable to game 21:20:30 ask people to join and game them 21:20:46 don't mention the mile wide holes that can be driven through this plan in IRC 21:20:49 *oops* 21:20:54 +1 21:21:00 a solid start to many things 21:21:07 * russellb points at reviewstats 21:21:20 also avoid the pathology of bad metrics that Ubuntu has seared in my consciousness 21:21:44 heh 21:22:27 k, well let's dig into it more next week 21:22:29 enjoy your holiday :) 21:22:31 #topic sub-teams 21:22:45 alright, each week we also hit on status of some sub-team efforts within nova (or the compute program in general) 21:22:56 one that i'd like to start checking in on regularly this cycle is novaclient 21:23:06 multiple people have expressed interest in helping, some of which i still need to follow up with 21:23:17 but melwitt is one, and she has stepped up to help bring some status info to our weekly meeting :-) 21:23:28 melwitt: i know we're just getting started, but any thoughts for today? 21:23:53 yes, I have gathered a few metrics for the novaclient bugs and reviews 21:23:58 awesome 21:24:37 there are 107 open bugs, i.e. bugs that are not status 'fix released'. 40 bugs in new status, 2 high importance bugs (I don't remember how to link LP bugs here) 21:25:11 there are 10 patch reviews up, all of which are being actively looked at/updated/reviewed, a couple of them are WIP 21:25:38 OK, so definitely need some help doing triage 21:25:38 you can just paste bug links 21:25:38 but sure we have a fancy bot in here for bug numbers ... 21:25:53 so the two High priority bugs, being looked at? 21:25:54 oh ok. https://bugs.launchpad.net/python-novaclient/+bug/1039572 21:26:02 let's try ... bug 1039572 21:26:14 no bug bot. 21:26:28 lp:1039572 21:26:28 nope 21:26:43 that first one looks fixed now 21:26:51 joe linked to a patch that has since been merged 21:26:55 the other is https://bugs.launchpad.net/python-novaclient/+bug/1116593 21:26:57 melwitt: do you have permissions to close bugs? 21:27:06 melwitt: you just need to join the nova-bugs team on launchpad 21:27:07 in theory 21:27:25 russellb: oh, yes I'm part of nova-bugs team 21:27:35 melwitt: basically, change "Confirmed" to "Fix Committed" 21:27:58 russellb: ah ok. thanks. I missed that it had been merged, I hadn't clicked on it 21:28:10 all good :) 21:28:13 that's why we're talking through them! 21:28:14 progress 21:28:25 ok, so the other one ... depends on a nova feature, and we have the nova feature as wishlist 21:28:31 so i think i'd change the novaclient bug to wishlist, as well 21:28:51 ok. will do that. 21:29:09 and like magic, no more High bugs :-p 21:29:14 Heh 21:29:21 thanks for bringing the info! 21:29:22 I never knew it could be so simple! thanks for the help 21:29:49 if anyone would like to help with novaclient metrics, or help clean up the novaclient bug queue, please talk to melwitt ! (and you all can talk to me, too) 21:29:56 sounds like the bug queue is the biggest area in need of cleanup 21:30:01 fyi: the magic that closes bugs in launchpad fails if there's a dot after the bug number. So tag maintainers need to look for bugs to manually mark fixed. 21:30:16 dripton: that seems like an oversight :) 21:30:29 in the magic 21:30:38 russellb: Can we form a subteam for third party compatibility efforts? I know atleast 5 folks who're interested 21:30:39 dripton: sounds like a bug needing fixing 21:30:48 russellb: there's a regex somewhere in the git magic that needs fixing. I'll look for it. 21:30:56 rohitk: that was next on my list! 21:30:58 dripton: great! 21:31:02 ok, next sub-team ... 21:31:08 we had a session on AWS (EC2) compatibility 21:31:13 dripton: probably openstack-infra/jeepyb ? 21:31:22 the outcome was basically that those interested needed to coordinate efforts in a subteam 21:31:28 rohitk: that was your session, right? 21:31:34 russelb: right 21:31:38 OK, great 21:31:52 so some thoughts on how to get started ... poll the openstack-dev mailing list for those that would like to join 21:31:59 aim to have at least a few people working together regularly 21:32:10 and then find a time that is reasonable convenient for everyone to meet once a week 21:32:15 and schedule an IRC meeting 21:32:32 and then hopefully one person out of that group can come here and give some status to the rest of the nova community 21:32:56 once you have a group, work to divide and conquer work, baby steps 21:33:07 increase some test coverage, close feature gaps, whatever you guys think is important 21:33:16 from a high level anyway :-) 21:33:21 russellb: sounds good, will get started on this 21:33:25 ok, awesome 21:33:31 please let me know if you want to discuss further or need any help 21:33:35 i'm happy to see this get going 21:33:43 russellb: absolutely, thanks 21:34:02 and i can try to attend the first meetings at least to help talk about how the work integrates back into nova 21:34:04 processes and such 21:34:46 one other thing that we need a sub-team going on is containers 21:34:52 we had a big discussion on the future of docker 21:35:02 and it sounds like we may go with a separate project completely 21:35:13 step 1 being come up with a straw man API for what such a project would expose 21:35:26 to help us really decide if it needs to be separate, or continue to be an extension to nova 21:35:45 samalba: around by chance? 21:36:06 i'm not sure we identified concretely who was going to run with this idea 21:36:06 but samalba did the docker driver :) 21:36:13 so i'm hoping he does! 21:36:40 for reference ... 21:36:42 #link https://etherpad.openstack.org/p/docker-nova-hkg 21:37:01 big list of people on there interested 21:37:03 russellb: I am here 21:37:13 samalba: just trying to identify next steps on our containers discussion 21:37:13 (reading the backlog) 21:37:30 samalba: are you (or someone) going to start organizing that? drawing up an API? 21:37:45 yes, I was planning to start the spec this week-end 21:37:50 perfect! 21:37:57 and involve the people from the etherpad early 21:38:03 sounds great, let me know how it goes 21:38:08 and i'll also try to check in with you 21:38:16 mainly draft the API + requirements for starting the implem 21:38:21 * russellb nods 21:38:26 sounds like a good place to start 21:38:43 I guess etherpad is the right place to draft 21:38:44 another key bit is some architecture ... specifically, the interaction with existing openstack services 21:38:52 exactly, it's included 21:38:53 such as, is it built on top of nova exclusively? 21:38:59 ok, so you're all over it :) 21:39:13 yeah, etherpad or wiki, whichever 21:39:26 or a git repo 21:39:27 russellb: do we want this to be part of the openstack compute project (not sure if I got terminology right there), as in we would have nova and x 21:39:28 yes, just did not have the chance to start yet... but it's planned for short term :-) 21:39:29 with just docs 21:39:38 ok 21:39:42 russellb: I guess I am jumping the gun, never mind 21:39:47 jog0: no, it's a good question 21:40:00 my opionion is that containers are a part of nova right now, so this effort should start as a sub-project of the compute program 21:40:26 program ooh thats the word I was looking for 21:40:27 if this moves forward far enough that it becomes a new project, we'll re-evaluate 21:40:27 programs are largely about groups of people 21:40:32 if the people working on it are largely separate from those doing nova, it would probably make sense to have a new program 21:40:40 if there's significant overlap, i would keep it under compute 21:40:49 from a technical standpoint, i think we could argue it either way 21:40:52 but it really comes down to the people 21:41:12 on a related note, every program is supposed to have a mission statement that helps capture the scope 21:41:15 i don't think we have one :) 21:41:21 "To not suck" 21:41:23 "compute" is our mission statement 21:41:24 :-p 21:41:26 mikal: and that 21:41:34 "To not suck at compute" 21:41:40 mikal: ++ 21:41:49 so anyway, let's start with an effort of the compute program to bootstrap 21:41:51 make sense? 21:42:07 russellb: I have to leave IRC for a bit, but I'll keep you in the loop for specing this containers service 21:42:13 samalba: sounds good! 21:42:16 ok, other subteams 21:42:21 scheduling, vmware, xenapi, hyper-v 21:42:28 anyone want to provide updates on those? 21:42:35 all had a session, or multiple sessions last week 21:42:38 scheduling - no meeting this week (recovering from the summit) 21:42:49 n0ano: sounds good, i'm still recovering too :) 21:43:05 lots of sessions from the summit, we'll start going over them next week. 21:43:10 sounds good 21:43:19 #topic blueprints 21:43:29 so we had a thread about blueprint process here 21:43:36 #link http://lists.openstack.org/pipermail/openstack-dev/2013-October/017290.html 21:43:51 I've started integrating our process tweaks into the main blueprints page 21:43:53 #link https://wiki.openstack.org/wiki/Blueprints 21:43:58 was just going to ask about that 21:44:01 if there is anything about the blueprint process not clear from that page, it needs love 21:44:08 (as of the last couple hours) 21:44:32 so, we need to start reviewing these things 21:44:34 #link https://blueprints.launchpad.net/nova/+milestone/icehouse-1 21:44:38 66 blueprints, all on icehouse-1 21:44:42 21:44:45 none on 2 or 3 21:44:55 i can make that 65 21:45:01 cool :) 21:45:12 if you have something that you don't think will be merged by 3 weeks from today, please go ahead and update 21:45:27 instance type / flavor rename is just a lot of busy work, so taking a while 21:45:29 i'd like to spend tomorrow starting a big push to catch up on blueprint reviews 21:45:47 so if you're on nova-drivers, or would like to help (anyone can help), please hang around #openstack-nova tomorrow 21:45:54 and we'll see how far we get 21:45:57 depending on how well it goes, we may have another day next week 21:46:36 any particular ones you guys want to talk about right now? 21:46:48 russellb: can you approve bp https://blueprints.launchpad.net/nova/+spec/pci-api-support? 21:46:55 heh 21:47:04 we are confident for icehourse-1 21:47:10 haha 21:47:13 we will be reviewing blueprints tomorrow :) 21:47:21 there are probably 50 that need to be looked at 21:47:26 ok 21:47:29 +1, I just added one :-) 21:47:31 so you're not alone 21:47:53 we just need to firm up our process, and then talk about it at the summit 21:47:57 and this week has been a bit of recovery 21:48:00 is there a review meeting, which everyone can join? 21:48:00 russellb: I like https://wiki.openstack.org/wiki/Blueprints#Blueprint_Review_Criteria 21:48:03 so i think we're finally to where we can catch up on this 21:48:21 shane-wan1: no specific time, just going to do it tomorrow on IRC 21:48:23 "tomorrow" 21:48:24 shane-wan1: please read the scroll back from like three minutes ago 21:48:26 jog0: cool thanks 21:48:42 ok, just disconnected, sorry 21:48:44 :D 21:48:55 let's start pasting this wiki link everywhere 21:49:14 a majority of people seem to screw up step 1 21:49:14 (targeting to a milestone) 21:49:14 targeting a milestone is a trigger to get review ... 21:49:17 russellb: that may help make it easier to review them tomorrow 21:49:17 and then i get emails "why isn't my thing reviewed yet" 21:49:19 because you missed step 1 21:49:36 jog0: yeah hope so, and we can tweak that criteria as we go 21:49:46 another thing .... while we have 66 on icehouse, we probably have a couple hundred others 21:49:59 i'd like to do a mass close of others after another week or so 21:50:05 next friday was the date i put on the ML a while back 21:50:06 ideally, icehouse-3 will be busy 21:50:18 i'll post a reminder 21:50:34 https://blueprints.launchpad.net/nova/ 21:50:34 358 total blueprints 21:50:44 need to identify what's actually active work, so close them all :) 21:50:44 oh, my god 21:50:49 and see what gets reopened i guess 21:50:55 close with a comment linking to process info 21:50:58 blueprints look like bugs 21:51:07 nova is busy :) 21:51:28 implement all the thigns 21:52:01 OK, I think that's all for now! 21:52:04 #topic open discussion 21:52:10 a few minutes for any other topics 21:52:26 I have a random topic 21:52:30 So, I'm approved for Utah 21:52:37 mikal: nice work 21:52:44 * beagles notes "fast work" 21:52:53 mikal: we decided it's only for people in the northern hemisphere though, sorry 21:53:04 mikal: for meetup, so quick? 21:53:37