21:02:15 <russellb> #startmeeting nova 21:02:16 <comstud> oh hai 21:02:17 <openstack> Meeting started Thu May 2 21:02:15 2013 UTC. The chair is russellb. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:18 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:20 <openstack> The meeting name has been set to 'nova' 21:02:24 <russellb> well hello nova friends! 21:02:33 <russellb> who's around to chat? 21:02:33 <alaski> hi! 21:02:33 <comstud> why hello there 21:02:37 <beagles> howdy 21:02:37 <cburgess> greetings 21:02:37 <senhuang> russellb: hello 21:02:38 <dripton> hi 21:02:38 <cyeoh> hi 21:02:39 <dansmith> <-- 21:02:39 <n0ano> o/ 21:02:41 <GheRivero> o/ 21:02:43 <comstud> o/ 21:02:50 <senhuang> o/ 21:02:50 <harlowja> yo 21:02:52 <beagles> or.. 21:03:00 <beagles> whatcha at 21:03:05 <russellb> nice, lots of people 21:03:15 <russellb> #link https://wiki.openstack.org/wiki/Meetings/Nova 21:03:18 <devananda> \o 21:03:22 <russellb> #topic blueprints 21:03:28 <maoy> me 21:03:30 <russellb> #link https://blueprints.launchpad.net/nova/havana 21:03:39 <russellb> We have 52 blueprints on the havana roadmap 21:03:41 <mikal> Hi 21:03:47 <russellb> nice work! thanks for everyone's help with getting this in shape 21:04:00 <russellb> is there anything known to be missing from this list that people here are planning to work on? 21:04:04 <timello> o/ 21:04:21 <mikal> russellb: the periodic tasks reworking is missing, I'll fix that 21:04:35 <russellb> mikal: ah yes, i had that on my list to ask you about 21:04:39 <senhuang> we are working on a group api extension 21:04:48 <russellb> senhuang: group scheduling stuff? 21:04:49 <devananda> russellb: i just realized there's not actualy a BP for the nova->ironic split. should there be (assuming that is approved)? 21:04:54 <russellb> senhuang: vm-ensembles is on there ... i think? 21:04:55 <harlowja> well the taskflow stuff that i was going to help drive is missing, its still being talked about though 21:05:00 <mikal> russellb: I should probably add deferred instance file delete too 21:05:00 <russellb> devananda: yes 21:05:06 <senhuang> rusellb: yes. gary, alex and i have discussed adding a new api extenstion to nova 21:05:09 <devananda> k. /me creates it 21:05:11 <russellb> devananda: maybe just call the nova blueprint deprecate-baremetal-driver? 21:05:18 <russellb> mikal: indeed 21:05:21 <comstud> 52... 3 per week! 21:05:26 <harlowja> :) 21:05:36 <harlowja> comstud u can do it all 21:05:37 <cburgess> mikal: Deffered instance file delete? 21:05:41 <comstud> no 21:05:44 <senhuang> russellb: this api extension will allow users/tenants request a group of instances and specify the relationship among them 21:05:51 <russellb> comstud: we implemented a bit over 60 for grizzly 21:06:10 <comstud> cool 21:06:11 <mikal> cburgess: mothballing and NFS instance storage both need to be able to defer deleting instance files 21:06:15 <russellb> senhuang: ok, so can we just update the description of https://blueprints.launchpad.net/nova/+spec/vm-ensembles for that? 21:06:25 <russellb> senhuang: also need an idea of when you guys think that will be ready for review/merging 21:06:28 <russellb> right now it's not on the havana list 21:06:34 <mikal> cburgess: http://www.stillhq.com/openstack/havana/000002.html (look for the deferred instance delete heading) 21:06:35 <senhuang> russellb: we are planning to have a separate blue print 21:06:52 <senhuang> russellb: will be ready next for review/comments 21:07:00 <russellb> senhuang: ok, well just ping me later once you have stuff ready. they need to be assigned and have a target milestone set, and then i'll put it in the havana plan 21:07:08 <hartsocks> Hey, sorry I'm late. 21:07:15 <russellb> goal was to have this as close to complete my this coming tuesday as we can 21:07:21 <senhuang> rusellb: will do. 21:07:56 <russellb> quick notes for new blueprints if you're filing them ... set an assignee, set a target milestone (havana-1, 2, 3), and propose for the 'havana' series, then it goes into my queue to review 21:07:57 <harlowja> russellb for the task management stuff which is still under a little flux are u thinking to keep that out of the offical list for now? 21:08:08 <russellb> harlowja: yeah, still too much in flux on the plan there i think 21:08:14 <harlowja> russellb sounds fine 21:08:16 <harlowja> thx 21:08:26 <russellb> for reference, we have 50 in the plan, but 175+ open 21:08:29 <harlowja> :) 21:08:32 <russellb> so lots still in discussion/planning/etc 21:08:43 <comstud> dansmith: looks good 21:08:44 <comstud> + if hasattr(self, get_attrname(name)): 21:08:44 <comstud> + primitive[name] = getattr(self, name) 21:08:48 <comstud> ^ bug there 21:08:49 <russellb> comstud: hey you! 21:08:50 <uvirtbot> comstud: Error: "bug" is not a valid command. 21:08:56 <comstud> oops 21:08:59 <dansmith> heh 21:09:02 <comstud> I thought this was window 4 21:09:06 <beagles> :) 21:09:07 <comstud> (continue) 21:09:10 <senhuang> rusellb: i think i am going to defer the cross projects (nova, cinder, quantum) blue print 21:09:19 <devananda> done: https://blueprints.launchpad.net/nova/+spec/deprecate-baremetal-driver 21:09:35 <russellb> senhuang: ok ... i think that one needs some more discussion / consensus building on the openstack-dev list 21:09:42 <russellb> senhuang: unassign from yourself if you're dropping it 21:09:59 <russellb> or we can close it if you want 21:10:31 <senhuang> russellb: what i mean is to continue the discussions on the topic, but maybe move the actual design/implementation later 21:10:33 <russellb> devananda: approved/updated 21:10:40 <russellb> senhuang: ah ok, that's fine 21:10:46 <russellb> senhuang: so the blueprint is fine where it is then 21:10:54 <senhuang> russellb: yes 21:11:02 <russellb> it's in the "we want to do this, but it's not clear how/when yet" state 21:11:17 <senhuang> russellb: you are 100% right. 21:11:22 <russellb> :) 21:11:46 <senhuang> russellb: it is better to add the group scheduling support now 21:11:55 <russellb> yeah that sounds good 21:11:59 <senhuang> russellb: which is pretty achievable within havana 21:12:06 <russellb> so the next step in blueprint tracking after this week will be switching over to tracking progress against havana-1 21:12:18 <russellb> we've done a great job populating the list, but havana-1 is way overloaded right now 21:12:30 <russellb> it doesn't look realistic to get all of it completed/reviewed/merged 21:12:37 <russellb> so some things are likely going to have to shift to havana-2 21:12:38 <maoy> who's leading the v3 API work? 21:12:55 <russellb> if you're assigned to havana-1 stuff, make sure it's realistic (going to be ready quite soon), or update to havana-2 please 21:12:59 <russellb> maoy: cyeoh 21:13:31 <russellb> #note Please file any remaining missing blueprints 21:13:46 <russellb> #note Please verify that your milestone targets are realistic. havana-1 is overloaded right now. 21:13:56 <russellb> anything else on blueprints? 21:14:29 <maoy> feel like we are missing a v3 API specification blueprint 21:14:42 <devananda> russellb: i'm going to pass 2 db-related BP's to boris-42, if you dont mind 21:14:45 <maoy> existing ones are implementations 21:14:51 <russellb> maoy: there is a top level v3-api blueprint ... but good point that i don't think there's a specific one for the spec 21:14:55 <devananda> doubt i will have time to focus on those with the baremetal work 21:15:06 <russellb> devananda: sure, sounds good. i saw some older ones assigned to you but wasn't sure on the status, so hadn't added them to the havana list 21:15:22 <russellb> makes sense 21:15:48 <devananda> russellb: yea, there are a few that no one is working on. i can unassign them 21:15:48 <russellb> #note may need another blueprint for the v3 API spec. maoy to follow up with cyeoh on this 21:15:54 <russellb> devananda: that would be good 21:16:04 <russellb> makes it more claer that they're open for taking 21:16:05 <johnthetubaguy> russellb: a few of the xenapi are likley of havana, just need to determine the milestone 21:16:28 <russellb> johnthetubaguy: sounds good! once a milestone is set, propose for the havana series and i'll take a look 21:16:41 <cyeoh> maoy: by specification you mean the produced documentation for all the extensions like we currently have for v2? 21:16:44 <johnthetubaguy> russellb: will do, off to bug the Citrix guys tomorrow 21:16:48 <russellb> :) 21:17:12 <russellb> cyeoh: something for api.openstack.org ... that kind of thing (is what i interpreted it as) 21:17:40 <russellb> but i guess it's not really ready to be published until we declare it "done" 21:17:43 <maoy> cyeoh: yeah things like http://docs.openstack.org/api/openstack-compute/2/content/ 21:17:45 <russellb> we need a period while it's in flux 21:17:56 <russellb> marked as beta/in progress/ or something 21:18:08 <johnthetubaguy> is that likely to extend beyond havana? 21:18:19 <cyeoh> russellb, maoy: ok, because of the way it is produced I was thinking it was part of the tests work, but I'll ad a separate blueprint for it. 21:18:26 <russellb> johnthetubaguy: ideally not IMO 21:19:00 <russellb> cyeoh: you may be right ... doesn't hurt to have a checklist item for "is it documented" at the end 21:19:06 <johnthetubaguy> russellb: cool, seems like a good first goal 21:19:32 <maoy> cyeoh: one of the problem i have with v2 API spec is related to the Error state. 21:19:35 <russellb> johnthetubaguy: thanks for joining us late :) 21:19:52 <russellb> ok, let's catch up on more v3 API details out of meeting 21:19:55 <russellb> #topic bugs 21:20:04 <maoy> sure 21:20:04 <russellb> So all this focus on new stuff has caused us to fall behind on bugs 21:20:07 <johnthetubaguy> russellb: no worries, just got back from a rehersal 21:20:15 <russellb> #link http://bit.ly/105Bydd 21:20:28 <mikal> 80 bugs! 21:20:33 <hemna> I'm working on another fibre channel issue we found while doing load testing 21:20:37 <russellb> I'm always open to ideas for ways to improve bug handling 21:20:43 <hemna> I should get that fixed tomorrow 21:20:45 <russellb> for now ... anyone up for a bug day? 21:20:51 <russellb> hemna: k 21:20:54 <mikal> A bug day sounds like a good idea to me 21:20:57 <russellb> we could do a bug day tomorrow ... 21:20:58 <beagles> o/ 21:21:06 <russellb> or next week if that's better for some reason 21:21:08 <mikal> Hmmm, can we do a day that's not Saturday for me? 21:21:11 * dansmith schedules vacation for tomorrow 21:21:14 <russellb> mikal: ha, yes 21:21:14 <senhuang> yep. i need a little bit more guidance on this bug https://bugs.launchpad.net/nova/+bug/1049249 21:21:16 <uvirtbot> Launchpad bug 1049249 in nova "Remove plugging of internal classes from configuration" [High,Confirmed] 21:21:17 <johnthetubaguy> +1 for bug day, maybe next week though 21:21:23 <cburgess> bug day sounds like fun. 21:21:27 <russellb> mikal: well really that would mean today is your bug day :) 21:21:27 <morganfainberg> +1 for bug day here 21:21:28 * cyeoh agrees with mikal ;-) 21:21:36 <morganfainberg> but yeah next week please ;) 21:21:40 <mikal> I propose... Wednesday! 21:21:48 <mikal> No one does work on Wednesdays anyways 21:21:49 <russellb> yeah, heads up would be nice 21:22:10 <russellb> wednesday is good for me 21:22:14 <russellb> cancel all your meetings, people! 21:22:18 <russellb> we have cleanup to do! 21:22:31 <senhuang> biweekly bug day? 21:22:32 <russellb> #note bug day proposed for Wednesday, May 8 21:22:50 <mikal> senhuang: having something regular is a good idea IMHO 21:22:55 <russellb> yeah i think so 21:23:01 <hemna> +1 21:23:02 <russellb> mikal: interested in organizing such a thing? 21:23:06 <mikal> Also, my timezone gets all the easy bugs... :P 21:23:09 <russellb> mikal: the one next week, and a regular one? 21:23:15 <mikal> russellb: sure, I shall send an email and so forth 21:23:18 <russellb> mikal: yay 21:23:29 <senhuang> mikal: great! 21:23:32 <russellb> mikal: and have stat tracking ready to help motivate? :-D 21:23:49 <russellb> my number is bigger than yours! etc 21:23:49 <mikal> russellb: sure 21:23:49 <cburgess> Biweekly or monthly 21:23:52 <johnthetubaguy> how regular? once a month? 21:23:55 <mikal> russellb: noting there might be bugs in the bug tracking 21:24:03 <dansmith> wednesdays are my biggest meeting days, 21:24:05 <russellb> mikal: patches welcome, right? 21:24:09 <russellb> dansmith: weak! 21:24:10 <dansmith> so cancelling them twice a month ain't gonna happen :) 21:24:13 <mikal> russellb: for sures 21:24:25 <dansmith> but I'm fine with bug days being on wednesdays otherwise :) 21:24:29 <russellb> dansmith: or is it perfect? you can work on bugs during meetings 21:24:30 <mikal> dansmith: what's your manager's email address? 21:24:37 <cburgess> LOL 21:24:37 <dansmith> haha 21:24:37 <mikal> Hah 21:24:39 <timello> heh 21:24:46 <morganfainberg> hehe 21:24:48 <beagles> is it weird that I like Monday's for that kind of thing? 21:25:03 <dansmith> beagles: yes, it's weird 21:25:05 <beagles> :) 21:25:20 * russellb will make any day work ... but please not Tuesday 21:25:21 <dripton> too many holidays on Mondays to make them good for regularly scheduled real work 21:25:29 <beagles> good point 21:25:36 <comstud> I vote for Tuesday 21:25:43 <russellb> comstud: figures 21:25:46 <morganfainberg> dripton: same problem with fridays… except also add in alcohol 21:25:48 <comstud> kidding 21:25:55 <hemna> mmm beer 21:25:59 <mikal> My Monday is special to me... Its my US free day where I get actual work done. 21:26:00 <comstud> Sunday 21:26:08 <morganfainberg> comstud: lol 21:26:12 <russellb> beer is acceptable during work on bug day 21:26:22 <russellb> with moderation, of course. 21:26:24 <comstud> Sunday is when I get most of my extra work done. 21:26:26 <beagles> everyday is bug day then? 21:26:26 <dripton> Just don't exceed the Ballmer Peak 21:26:46 <mikal> Anyways, moving on... 21:26:46 <mikal> :P 21:26:48 * beagles points to the beer=bug day correlation 21:26:49 <hartsocks> beagles: I honestly thought everyday *was* bug day… 21:26:54 <russellb> #topic mikal going to work on organizing a regular scheduled bug day so we stay on track 21:27:19 <russellb> well, people get wrapped up in new development, or day job requirements 21:27:43 <russellb> so "bug day" is trying to get everyone to set other things aside, all at the same time, so we can have some focused time together improving our bugs situation (triaging, fixing, etc) 21:27:50 <russellb> that's my take anyway 21:28:03 <mikal> Agreed 21:28:07 <johnthetubaguy> +1 21:28:07 <hartsocks> russellb: thanks for explaining to the newb. 21:28:08 <russellb> #undo 21:28:08 <openstack> Removing item from minutes: <ircmeeting.items.Topic object at 0x295c390> 21:28:21 <russellb> #note mikal going to work on organizing a regular scheduled bug day so we stay on track 21:28:24 <russellb> note, not topic! 21:28:40 <russellb> mikal: anything else on bugs? 21:28:42 <russellb> or anyone else 21:28:45 <senhuang> could you roll back? :P 21:29:03 <russellb> senhuang: yeah i did #undo, it just doesn't actually fix the IRC channel topic, but it will fix the minutes 21:29:12 <harlowja> russellb sounds good, lets make it happen 21:29:15 <mikal> russellb: nup, that's me 21:29:18 <russellb> k 21:29:21 <russellb> #topic open discussion 21:29:30 <dansmith> I have something for open discussion 21:29:32 <russellb> some open discussion time, then we'll close out 21:29:37 <russellb> dansmith: go! 21:29:51 <dansmith> despite comstud's best efforts, he did not paste a dansmith-authored bug into the middle of the meeting 21:29:53 <dansmith> that is all. 21:29:59 <harlowja> :) 21:29:59 <comstud> LOL 21:30:05 * russellb slow claps 21:30:07 * beagles snickers 21:30:09 <maoy> haha 21:30:12 <comstud> dansmith is correct. 21:30:21 <comstud> Sorry about that. 21:30:35 <mikal> Heh 21:30:47 <dansmith> okay sorry, real open discussion now :) 21:31:22 <comstud> I am so used to finding dansmith-authored bugs that it just becomes too natural. 21:31:28 <dansmith> ouch :( 21:31:41 <comstud> I kid, I kid. 21:31:44 <russellb> open discussion is also the one time of week where it's acceptable to shamelessly plug your review 21:31:44 <comstud> I just cannot read python. 21:31:51 <russellb> comstud: ORLY 21:32:04 <comstud> Yes 21:32:26 <comstud> that's it. i'm rewriting this all in C. 21:32:35 <russellb> random thing i've been thinking about ... i'd like to start a wiki page that's a Nova Review Checklist 21:32:41 <morganfainberg> comstud: i was expecting erlang. 21:32:45 <lifeless> comstud: cool. 21:32:46 <russellb> so we can start keeping a central list of things all reviewers should be looking for 21:32:56 <morganfainberg> russellb: that would be awesome 21:33:06 <cyeoh> my shameless plug then: https://review.openstack.org/#/c/27276/ (first of the v3 API extension framework patches) 21:33:13 <russellb> cyeoh: ooh good plug 21:33:14 <comstud> russellb: That would be helpful. 21:33:21 <johnthetubaguy> sounds good, like the RPC versioning 21:33:22 <russellb> comstud: yeah, figure it could help with consistency 21:33:27 <comstud> I often find myself digging way down intot he list of reviews to try to find the most important ones 21:33:59 <comstud> (not that I shouldn't be reviewing ALL OF THE THINGS, but) 21:34:24 <comstud> sometimes I only have time to knock out a few.. and I want to make sure I'm hitting the high priority ones 21:34:28 <russellb> ha, apparently i had this idea a long time ago, but never really completed it 21:34:32 <russellb> https://wiki.openstack.org/wiki/ReviewChecklist 21:34:40 <beagles> yeah, some conventional wisdom would be cool.. great to point to for newbs, etc. 21:34:44 <russellb> so, i'll try to start making that more useful, and post to openstack-dev when it's worth reading 21:34:51 <dansmith> yay 21:34:58 <russellb> others certainly welcome to help get it going! 21:35:14 <russellb> #note let's work on https://wiki.openstack.org/wiki/ReviewChecklist so that we have a more complete nova review checklist to help with review consistency! 21:35:21 <beagles> got one item.. if you are not core and you notice that a review has been reviewed by 5 other non-core guys, find a core guy and tug their shirt-sleeve 21:35:33 <senhuang> russellb: that will be awesome for new developers like me 21:35:56 <russellb> #note It will also help communicate expectations to patch submitters 21:36:07 <beagles> (a note for your wiki, that is) 21:36:15 <russellb> beagles: edit away! 21:36:20 <timello> that's interesting, specially for new reviewers and also for experienced ones... will make reviews more consistent 21:36:26 <beagles> yup 21:37:29 <russellb> alrighty, any other topics? 21:37:58 <johnthetubaguy> not today, might try roll up stuff from xenapi meeting next week, if there is anything pressing 21:38:06 <harlowja> just throwing this out there, for others to read the orchestration weekly is starting back up 21:38:13 <harlowja> #link http://eavesdrop.openstack.org/meetings/openstack_orchestration_group/2013/ 21:38:17 <russellb> i should probably start adding an agenda item for subteam reports 21:38:19 <harlowja> lots of good stuff i think occuring there :) 21:38:31 <harlowja> russellb that'd be useful i think 21:38:39 <johnthetubaguy> russellb: +1 21:38:39 <comstud> russellb: what's status of scheduler filters/weights -> entrypooints ? we just dropping it completely for now? 21:39:02 <n0ano> anyone interested in the scheduling sub-group we're just starting, the last irc log is here: http://eavesdrop.openstack.org/meetings/scheduler/2013/scheduler.2013-04-30-15.04.log.html 21:39:03 <russellb> comstud: needs to get done ... nobody is doing it 21:39:07 <russellb> comstud: deadlock 21:39:18 <comstud> russellb: that works out well for me right now 21:39:22 <comstud> but was curious 21:39:23 <comstud> :) 21:39:30 <russellb> yeah, i don't see it on the radar 21:39:32 <comstud> Ok 21:39:38 <russellb> but good point, i'll put that on my list of "things that need volunteers" 21:39:42 <comstud> I got the cells filt/weights coming up soon 21:39:49 <comstud> just adding a test or 3 21:39:53 <russellb> i'm going to do a big post soon on "what's in our havana plan" and "things we need help with" 21:40:00 <senhuang> anyone interested in the idea of having scheduling make decision based on some external information? 21:40:03 <harlowja> russellb cool 21:40:26 <russellb> senhuang: external? like the weather? 21:40:32 * russellb kids 21:40:36 <comstud> russellb,senhuang: reminds me. i wanted to create a BP for conductor/scheduler changes.. that we should start with before other larger scheduler changes 21:40:41 <comstud> and bring it up on list 21:40:49 <russellb> comstud: make it happen! 21:40:49 <senhuang> russellb: like some static topology configuration 21:41:06 <alaski> comstud: I made a blueprint for that, but didn't bring it up to the list yet 21:41:07 <morganfainberg> senhuang: you mean something like a constraint system? 21:41:18 <russellb> senhuang: yeah. vishy seemed interested in that at the summit 21:41:24 <comstud> alaski: Oh yeah, that's right. I forgot that was going to be a *you* thing 21:41:27 <comstud> GOOD 21:41:29 <comstud> :) 21:41:33 <russellb> which thing 21:41:33 <comstud> I have too much already 21:41:36 <alaski> https://blueprints.launchpad.net/nova/+spec/query-scheduler 21:41:39 <russellb> query scheduler? 21:41:40 <russellb> ha 21:41:41 <comstud> russellb: conductor/scheduler change 21:41:43 <russellb> hadn't even seen the blueprint 21:41:44 <alaski> just targeted it 21:41:45 <russellb> called it! 21:41:52 <harlowja> alaski +1 21:41:56 <harlowja> def need that i think 21:42:01 <russellb> +1 21:42:03 <russellb> will approve 21:42:04 <senhuang> morganfainberg: not sure. it is something like physical connectivity information put on some files. schedulers can read this information from files 21:42:09 <russellb> alaski: havana-1 realistic? 21:42:33 <harlowja> alaski if u need any help with that one let me know 21:42:42 <alaski> I guess I should find out exactly when that is, but I don't think it will be too large 21:42:51 <russellb> alaski: toward the end of May 21:42:55 <alaski> harlowja: cool 21:43:00 <morganfainberg> senhuang: ah yeah, similar in thought (aka, schedule X instances but lay it out in Y configuration) vs. just resource allocation/simple filtering 21:43:08 <comstud> I think it will be relatively small work. new sched method(s) so that backwards compat works 21:43:08 <russellb> alaski: https://wiki.openstack.org/wiki/Havana_Release_Schedule 21:43:10 <russellb> May 30 21:43:13 <alaski> russellb: I think so, keeping it pretty simple 21:43:29 <russellb> k approved for havana 21:43:30 <morganfainberg> senhuang: and base it on your "other" information. 21:43:41 <morganfainberg> senhuang: that could be useful. 21:43:50 <senhuang> moganfainberg: yep. then the scheduler is not limited to the information provided by host state reported by host-manager 21:44:02 <morganfainberg> senhuang: exactly. 21:44:09 <russellb> alaski: thanks! 21:44:50 <senhuang> morganfainberg: maybe we should start thinking of writing a blue print and see how to extend the current scheduler module 21:44:50 <alaski> np. This should make shelving a bit easier as well 21:45:37 <russellb> alaski: which we also have on havana-1 21:45:43 <russellb> alaski: you have work to do! :) 21:46:05 * alaski types faster 21:46:24 <alaski> shelving may get pushed, but I'm hopeful 21:46:30 <russellb> k 21:46:34 <morganfainberg> senhuang: i wouldn't be opposed to helping with something like that. I'm wrangling cburgess and a couple other people here so we can get some blueprints in asap so it can all get done for havana, so i'll be in BP mode soon. 21:46:59 <russellb> alaski: we could push, and always pull it back in if things go well 21:47:20 <senhuang> morganfainberg: that will be great. i think some kind of wrapper around hoststate could be possible approach 21:47:26 <alaski> russellb: works for me, I'll change it 21:47:29 <russellb> k 21:48:11 <russellb> alright, it was a pleasure talking to you all 21:48:12 <russellb> thanks! 21:48:15 <russellb> #endmeeting