21:02:25 <russellb> #startmeeting nova 21:02:26 <openstack> Meeting started Thu May 23 21:02:25 2013 UTC. The chair is russellb. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:27 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:30 <russellb> hello, everyone 21:02:30 <openstack> The meeting name has been set to 'nova' 21:02:32 <devananda> \o 21:02:34 <russellb> who's around? 21:02:35 <alaski> o/ 21:02:37 <cyeoh> hi! 21:02:40 <morganfainberg> oh hi! 21:02:42 <mriedem> hi 21:02:45 <russellb> #link https://wiki.openstack.org/wiki/Meetings/Nova 21:02:48 <n0ano> o/ 21:02:50 <beagles> .o/ 21:03:15 <senhuang> o. 21:03:17 <russellb> #topic havana-1 21:03:23 <harlowja> yo 21:03:27 <russellb> #link https://launchpad.net/nova/+milestone/havana-1 21:03:28 <dansmith> o/ 21:03:33 <russellb> soooo, havana-1 is a week from today 21:03:40 <russellb> the feature merge deadline is Tuesday 21:03:41 <mikal> Yay! 21:03:52 <russellb> and there are a bunch of things needing review to make that happen 21:04:16 <dansmith> the apiv3 stuff needs major review 21:04:20 <russellb> so, let's try to make the things on this list a priority with your review time 21:04:30 <russellb> #help APIv3 work needs reviews: http://goo.gl/5kChi 21:04:58 <russellb> #help Please use havana-1 blueprints to prioritize your review time in the next few days 21:05:41 <russellb> any other comments or questions about the push toward havana-1? 21:05:58 <russellb> also a bunch of bugs targeted at havana-1, but all in progress (presumably under review) 21:06:03 <dansmith> objects! 21:06:09 <mikal> Don't forget h-1 targetted bugs either 21:06:13 <russellb> mikal: :-) 21:06:15 <morganfainberg> dansmith: objects! 21:06:19 <mriedem> how does one get a bug targeted for h-1? 21:06:20 <mikal> (Curse my slow typing) 21:06:27 <russellb> dansmith: orly? would you like to say more? :-) 21:06:49 <russellb> mriedem: not sure everyone has permissions to do it 21:06:53 <dansmith> russellb: I'm on a call and trying to make points of things to review.. apiv3 and objects, both targetted for h1 21:06:55 <russellb> mriedem: you can ask nicely :-) 21:07:02 <russellb> ah 21:07:13 <russellb> #help please help review dansmith's objects patches, too 21:07:13 <mriedem> russellb: ok, i'll remember in open topics, thanks 21:07:31 <russellb> alright, so yeah, let's get this stuff up 21:07:39 <russellb> we have 3 dev milestones to get features in 21:07:47 <russellb> and right now the havana-1 list is about 1/3 of what we have on the full havana list 21:07:57 <russellb> so we really do need to get these done to stay on track 21:08:01 <devananda> boris is asking for reviews on the bp/db-improve-archiving reviews, too 21:08:15 <russellb> #help please help review db-improve-archiving patches 21:08:30 <russellb> k, on to some bug discussion 21:08:31 <russellb> #topic bugs 21:08:38 <russellb> we had a bug day yesterday ... good news and bad news 21:08:45 <russellb> good news: a bunch of bugs got triaged 21:08:51 <russellb> bad news: it was basically all by one person 21:08:57 <russellb> huge gold star for mikal :-) 21:09:16 <russellb> so, mikal got a thread going on the mailing list trying to discuss new ideas for handling our triage workload 21:09:28 <russellb> my favorite idea so far is something quantum is doing, which is tagging 21:09:39 <mikal> So, having now read that thread... I like the tagging idea. Let's try it. 21:09:41 <russellb> and then having people focused on triaging bugs with a given tag that they know well enough 21:09:57 <russellb> mikal: cool 21:10:02 <mikal> It becomes a nice way of getting a subteam's attention as well 21:10:05 <russellb> so, i started documenting this to get us going 21:10:05 <mriedem> russellb: mikal: would work well for the virt drivers, they already have tags 21:10:11 <russellb> #link https://wiki.openstack.org/wiki/Nova/BugTriage 21:10:13 <russellb> mriedem: yeah 21:10:28 * mikal wonders if we need a catchall tag as well? 21:10:30 <russellb> so, take a look at that wiki page ... we need to flesh out our tag list, and start getting people assigned to each tag 21:10:35 <devananda> i haevn't read the thread, but, tagging bugs eg "db", "vmware", etc seems really helpful to me 21:10:35 <mriedem> right, i try to tag all powervm bugs and triage them (if not already), but i've noticed it doesn't happen often 21:10:39 <russellb> mikal: "core stuff" ? heh 21:10:40 <mriedem> agreed 21:10:50 <mikal> russellb: "unknown"? 21:10:57 <mikal> For "I'm not sure WTF this is, but it sure is a thing" 21:11:00 <russellb> the other thing is we need some canned queries so people can quickly pull up the list applicable to them 21:11:08 <devananda> is there a good way to encourage bug submitters to add tags? 21:11:13 <russellb> tagged with db and in the new state, or whatever 21:11:27 * comstud is here now 21:11:28 <russellb> devananda: i don't know of a way to have a blatantly obvious bug guidelines thing 21:11:30 <mikal> Could we use triaged and confirmed here? 21:11:38 <mikal> Mark things triaged when tags added / verified 21:11:42 * johnthetubaguy is here now too 21:11:46 <mikal> Mark with confirmed when someone on that subteam has evaluated the bug? 21:11:52 <mriedem> russellb: the hypervisor support matrix wiki has queries for known virt driver bugs 21:12:01 <russellb> mikal: if needed ... seems like the lack of a tag or has a tag is enough 21:12:13 <mriedem> https://wiki.openstack.org/wiki/HypervisorSupportMatrix#Known_bugs.2C_by_hypervisor 21:12:14 <russellb> mriedem: indeed ... just need to refine the queries to show you the tagged bugs that are untriaged 21:12:21 <mriedem> yeah 21:12:22 <mikal> russellb: well, wont the lack of a tag keep the bug stuck in triage hell? 21:12:25 <russellb> and then get those on the new bug triage page 21:12:39 <russellb> mikal: i was hoping we'd have another query for untriaged bugs that are untagged 21:12:44 <russellb> mikal: and people looking at that list, as well 21:12:44 <johnthetubaguy> I tend to google xen in the bugs and tag a few 21:12:57 <johnthetubaguy> (and triage them) 21:13:02 <mriedem> johnthetubaguy: that's what i did for powervm the other night 21:13:06 <russellb> johnthetubaguy: awesome, so basically already doing this :-) 21:13:17 <devananda> mikal: that's my concern. i've spotted several baremetal bugs come in w/o a tag 21:13:21 <russellb> yeah, so we need to formalize this, and do it consistently, make it easy, etc 21:13:27 <devananda> found them by actually using search 21:13:33 <johnthetubaguy> russellb: +1 21:13:54 <mikal> We can always iterate 21:13:55 <russellb> devananda: yep, so i'm proposing we have a first step of triage that is just tagging, before the "real" triage happens 21:13:58 <mikal> Let's give it a try and see what happens 21:14:01 <johnthetubaguy> devanada: yes, theres always a few that don't get the tag, users probably don't know the tags, and thats fine 21:14:02 <mikal> It can't be worse than now 21:14:18 <russellb> so take a look at https://wiki.openstack.org/wiki/Nova/BugTriage and help fill it in :-) 21:14:31 <johnthetubaguy> sounds like a good plan, I am happy to take XenAPI stuff 21:14:36 <russellb> johnthetubaguy: perfect 21:14:49 <mikal> I'll add myself to the list of people dumb enough to tag things 21:15:19 <russellb> mikal: cool :-) ... i added myself because i thought it'd be a good spot in the process to quickly see what bugs are coming in, but without having the full responsibility of full triage anymore 21:15:25 <russellb> mikal: a nice breakdown, i think 21:15:50 <mikal> russellb: yeah, there's also a big backlog of untagged bugs, so we might have to grind through them for a bit 21:16:00 <cyeoh> is there a way to show bugs that don't have a tag at all? 21:16:19 <russellb> mikal: indeed ... if we do it in the triage part, everything new will always be tagged ... but would be good to fix the older stuff over time 21:16:26 <cyeoh> (and I'm happy to triage bugs that are REST API related) 21:16:28 <russellb> cyeoh: don't know, i was hoping we could have a query for that 21:16:43 <russellb> i'd like to query for new bugs without a tag, that'd be the query used for step 1 on https://wiki.openstack.org/wiki/Nova/BugTriage 21:16:50 <russellb> cyeoh: consider yourself signed up :-) 21:17:44 <russellb> i don't know if you can do a query for that ... but i haven't tried yet 21:17:55 <russellb> and i think we're a pretty smart group, so i bet we can figure something out. 21:17:59 <russellb> even if we script it. 21:18:26 <russellb> cool ... so at least we have a plan for a next thing to try on bug triage 21:18:31 <russellb> so hopefully we won't make mikal sad again 21:18:50 <mikal> Oh, I am sure we will 21:18:53 <russellb> :-) 21:18:58 <mikal> Just in new and exciting ways 21:18:58 <mikal> :P 21:19:02 <russellb> yep 21:19:10 <russellb> but at least not in the same way 21:19:13 <russellb> see, that's progress 21:19:38 <russellb> anything else on bugs? 21:19:42 <cyeoh> ah, apparently searching for '-*' in the tags field searches for bugs without tags 21:19:56 <russellb> cyeoh: nice :-) 21:20:09 <russellb> #topic sub-team reports 21:20:18 <russellb> who's around to give a subteam report? raise your virtual hand and we'll go one by one 21:20:24 * johnthetubaguy waves 21:20:24 <russellb> devananda: want to start? 21:20:29 <n0ano> scheduler o/ 21:21:01 <russellb> ok, will come back to devananda 21:21:04 <russellb> johnthetubaguy: xen! 21:21:09 <johnthetubaguy> hey 21:21:27 <johnthetubaguy> major news, smokestack is getting closer mate and dan price are working on it 21:21:37 <russellb> great to hear 21:21:45 <johnthetubaguy> on latest XenServer 6.1 rather than 5.6 21:21:53 <johnthetubaguy> thanks to Ant from Rax for that 21:21:59 <johnthetubaguy> other stuff is boring really, so that everything 21:22:05 <russellb> k, thanks :-) 21:22:06 <devananda> back now. also, dripton, are you around? 21:22:09 <dripton> hi 21:22:14 <russellb> devananda: alright, go ahead :-) 21:22:16 <antonym> johnthetubaguy: no prob :P 21:22:17 <dripton> I can give one for db though we're not strictly a nova team 21:22:33 * devananda passes the db-mic to dripton 21:22:39 <russellb> dripton: that's fine, other teams are like that too 21:22:39 <harlowja> i can give something to 21:22:47 * dripton takes mic 21:22:49 <russellb> harlowja: ack, you're in the queue 21:23:14 <dripton> So boris-42 did a whole bunch of patches for the improve-db-archiving blueprint, which are in good shape but several await +2s 21:23:27 <dripton> We'll be nagging reviewers before the havana-1 deadline to get those in. 21:23:50 <dripton> Also comstud is making a big change for the mysqldb backend, but that won't be ready for a bit longer. 21:23:58 <dripton> And various small fixes. 21:24:00 <dripton> That's it. 21:24:07 <russellb> cool, thanks! 21:24:11 <russellb> devananda: Ironic? 21:24:15 <devananda> sure 21:24:21 <devananda> first, quick note about nova-baremetal 21:24:27 <devananda> there are two bugs tagged critical 21:24:34 <devananda> looks like one is being worked in (i need to review after this) 21:24:47 <devananda> this one may need some love 21:24:48 <devananda> https://bugs.launchpad.net/nova/+bug/1178092 21:24:49 <uvirtbot> Launchpad bug 1178092 in nova "second boot during baremetal deploy does not configure netboot : will hang unless the machine attempts PXE automatically" [Critical,Triaged] 21:24:58 <devananda> unless someone else gets to it first, i will try this weekend 21:25:07 <devananda> ironic ... 21:25:19 <devananda> lots of people came out of the wood work and we've divvied up tasks 21:25:21 <devananda> so it's moving :) 21:25:25 <devananda> nothing for h1 tho 21:25:28 <devananda> [eol] 21:25:31 <russellb> oh cool, was hoping people would step up 21:25:47 <russellb> glad to hear 21:25:57 <devananda> ah, i suppose this is helpful too 21:25:57 <russellb> devananda: thanks! 21:26:03 <devananda> #link https://blueprints.launchpad.net/ironic 21:26:04 <devananda> :) 21:26:08 <russellb> nice 21:26:12 <russellb> getting all official 21:26:15 <devananda> hehe 21:26:22 <harlowja> wow, i need that, lol 21:26:31 <russellb> k, couple more teams 21:26:35 <russellb> n0ano: scheduler! 21:26:38 <n0ano> Got through two items, network bandwidht scheduling & group/ensembles, mostly discussion over mechanics of how to make the changes. 21:26:50 <n0ano> senhuang, was there the whole time, maybe he can add more 21:27:28 <n0ano> otherwise that's about it, go through the log for the gory details 21:27:30 <senhuang> Yep. there were a lot of discussions on both subjects 21:27:46 <n0ano> #link http://eavesdrop.openstack.org/meetings/scheduler_sub_group/2013/scheduler_sub_group.2013-05-21-15.00.log.html 21:27:48 <senhuang> at the end of the meeting, people also discussed the multi-scheduling 21:28:10 <russellb> i'm pretty sure i had a dream this week involving multi-scheduling, it was really bizarre. 21:28:27 <senhuang> we are making progresses towards the instance group api extension implementation 21:28:39 <senhuang> db part of work is done 21:28:59 <russellb> so, I put a -2 on the db part, but just because i'd like to see more of the code that uses it before it goes in 21:29:08 <senhuang> now the api part of work is submitted, but there is still some work on the REST part to submit 21:29:25 <russellb> oh, didn't see the next update, then 21:29:38 <senhuang> this work uses the db part of work 21:29:44 <russellb> cool 21:29:44 <senhuang> i hope i can finish it this week 21:30:05 <senhuang> okay. done 21:30:15 <russellb> thanks! 21:30:21 <russellb> harlowja: you're up 21:30:24 <harlowja> sweet! 21:30:45 <harlowja> so library is still being worked on, lots of good work there, forming up quite nicely 21:31:05 <harlowja> working with johnthetubaguy and devananda and others on various pieces of it 21:31:12 <harlowja> discussing how lock ownership/service might work 21:31:21 <harlowja> prototyping how cinder might use it 21:31:49 <harlowja> database tracking backend got hooked in, so good progress going 21:31:54 <harlowja> *moving to stackforge shortly* 21:32:04 <harlowja> #link https://review.openstack.org/#/c/29981/ 21:32:08 <harlowja> thats about it :) 21:32:13 <russellb> cool, thanks! 21:32:15 <harlowja> np 21:32:25 <russellb> hartsocks: hi! what's up in vmware land 21:32:29 <hartsocks> hey 21:32:35 <hartsocks> #link http://eavesdrop.openstack.org/meetings/vmwareapi/2013/vmwareapi.2013-05-22-17.01.html 21:32:43 <hartsocks> So we met this week with 11 participants. 21:32:52 <hartsocks> We're targeting H-2 for two of our blueprints. 21:33:06 <hartsocks> And we identified 4 "blocker bugs" that need to get fixed ASAP 21:33:14 <hartsocks> We have folks working on all 4 of these. 21:33:26 <hartsocks> So I'd say things are starting to really get rolling for us. 21:33:28 <russellb> k, ping me if you think any can get merged in the next few days 21:33:45 <hartsocks> I think we need to get some more disucussion, but thanks. 21:34:03 <russellb> i meant those blocker bugs 21:34:06 <hartsocks> Right. 21:34:26 <russellb> but cool, i think i owe you guys blueprint review on the havana-2 blueprints 21:34:33 <russellb> haven't gone back to look to see if they've been updated 21:35:01 <hartsocks> We'll ping you when things move then. 21:35:05 <russellb> k 21:35:07 <russellb> anything else? 21:35:18 <hartsocks> That's about all. 21:35:23 <russellb> any questions on any sub-team stuff? 21:35:59 <harlowja> not currently, likely will be after i start jumping back into nova code :) 21:36:14 * russellb nods 21:36:19 <russellb> #topic open discussion 21:36:41 <russellb> mriedem: you had a bug? 21:36:51 <mriedem> yeah, i'm looking to get some reviews on this: https://review.openstack.org/#/c/29005/ 21:37:07 <mriedem> the changes are pervasive but not that bad, the size is mostly in tests 21:37:22 <russellb> yeah, diffstat looks scary 21:37:24 <mriedem> dansmith and comstud handled a lot of the low hanging review early 21:37:29 <russellb> (touching all the drivers) 21:37:35 <mriedem> yup 21:37:38 <mriedem> latent bug 21:38:02 <mriedem> i was also hoping to get this marked for h-1, i think it's pretty much ready to go: https://review.openstack.org/#/c/27534/ 21:38:10 <mriedem> mikal: ^ 21:38:30 <russellb> on the first one, based on the commit message, it makes sense to me 21:38:31 * mikal looks 21:38:34 <russellb> just needs careful review 21:38:45 <mikal> Oh yeah, that's the one where I was being a grumpy old man 21:38:57 <russellb> mikal: that's different? 21:39:02 <mikal> Herh 21:39:02 * russellb snickers 21:39:19 <mriedem> mikal: i think he took care of what you wanted, i.e. opening the perf bug 21:39:28 <mikal> Yeah he did 21:39:35 <mikal> Then I got distracted by a shiney object 21:39:39 <mikal> Reviewing it now 21:39:45 <mriedem> that will happen :) 21:39:48 <mriedem> thanks 21:40:23 <russellb> oh, so random thing ... i was playing with stats earlier. We've had 137 people participate in nova code reviews in the last 30 days 21:40:27 <russellb> i think that's pretty amazing. 21:40:40 <harlowja> def, where are all these people ;0 21:40:48 <dansmith> not as amazing as 138 would be, but definitely awesome :) 21:40:54 <mikal> LOL 21:40:57 <russellb> harlowja: scattered around the globe, i suspect 21:40:58 <mikal> dansmith: you're so upbeat 21:41:01 <dansmith> heh 21:41:03 <russellb> dansmith: ha, so hard to please 21:41:11 <dripton> you mean you don't each have a couple of sockpuppet to give your patches +1s? 21:41:13 <mikal> I'd like more code review friends in the southern hemisphere 21:41:22 <mikal> We might upduct dansmith and bring him down here 21:41:22 <harlowja> dripton, hahaha 21:41:31 <dansmith> heh 21:41:33 <mikal> s/upduct/abduct/ 21:42:44 <russellb> mikal: you should just move this way 21:42:57 <russellb> mikal: i can offer you sweet potato casserole as a relocation bonus 21:43:04 <mikal> russellb: but the bread is weird in 'merica 21:43:13 <mikal> russellb: you made that up. That's not a thing! 21:43:42 <russellb> mikal: with marshmallow. 21:43:50 <dansmith> mmmm, reminds me of home in the south 21:43:52 <mikal> Does arguing about food indicate that this meeting is kindof over? 21:43:59 <russellb> mikal: yes 21:44:01 <russellb> anything else? 21:44:12 <mikal> That sweet potatoe marshmallow thing is awesome though 21:44:19 <mikal> We should have a meetup in Hawaii 21:44:30 <russellb> alrighty then ... #openstack-nova is open 24/7 for other spontaneous discussion throughout the week 21:44:32 <cyeoh> +1 for Hawaii meetup! 21:44:35 <russellb> thanks all! 21:44:38 <russellb> cyeoh: mikal +100 21:44:41 <mikal> Laters! 21:44:52 <russellb> #endmeeting