21:02:00 <russellb> #startmeeting nova
21:02:01 <openstack> Meeting started Thu Oct 17 21:02:00 2013 UTC and is due to finish in 60 minutes.  The chair is russellb. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:02:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:02:05 <openstack> The meeting name has been set to 'nova'
21:02:09 <russellb> hello, everyone!
21:02:13 <mriedem> hi
21:02:15 <cyeoh> hi!
21:02:17 <Guest44600> hi.
21:02:20 <tjones> hi
21:02:22 <alaski> hi
21:02:29 <russellb> congratulations on the havana release!  \o/
21:02:35 <dansmith> woo
21:02:42 <dripton> yay
21:02:42 <jog0> !!
21:02:43 <openstack> jog0: Error: "!" is not a valid command.
21:02:49 <russellb> i considered canceling today ... as dansmith put it, "today should be a day where the class watches a movie while the teacher sleeps off her hangover"
21:03:01 <dansmith> heh
21:03:05 <russellb> i was amused
21:03:09 <tjones> lol
21:03:12 <russellb> but we can catch up quickly :)
21:03:24 <russellb> so things going on ... design summit!
21:03:29 <russellb> session proposal deadline is today
21:03:38 <russellb> and starting tomorrow we will be deciding on the session list and making a schedule
21:03:47 <russellb> hopefully will have the draft schedule completed by the end of next week
21:04:25 <russellb> Icehouse blueprints - please file them :-)
21:04:38 <russellb> to get them reviewed, the trigger is to target them to a release milestone (icehouse-1/2/3)
21:05:01 <russellb> at some point (probably next month) we're going to go through and close most things untargeted as effectively abandoned
21:05:04 <russellb> to clean up the list a bit
21:05:23 <russellb> i'll post to the ML about that, as well
21:05:30 <tjones> russellb: other that adding the proposal - here http://summit.openstack.org/cfp/details/246 - is there anything else i need to do to to bring it to your attention for review tomorrow?
21:05:38 * russellb looks
21:05:50 <tjones> oops - wait
21:06:05 <russellb> link isn't working
21:06:16 <tjones> yeah - just a sec *blush*
21:06:33 <russellb> http://summit.openstack.org/cfp/details/247 ?
21:06:53 <tjones> yes
21:07:01 <russellb> looks good
21:07:04 <tjones> thanks
21:07:08 <russellb> np
21:07:15 <russellb> any other blueprint questions?
21:07:25 <russellb> err i meant summit
21:07:28 <russellb> but summit or blueprint i guess
21:07:46 <russellb> next thing ... we've been talking a lot about CI for each compute driver
21:07:54 <russellb> congrats to the VMWare team for getting theirs up and running!
21:08:06 <dansmith> yeah!
21:08:11 <russellb> you will now start seeing "VMWare Mine Sweeper" vote on reviews
21:08:20 <jog0> very exciting
21:08:40 <russellb> tjones: really, totally thrilled to see the progress, setting a good example for others
21:08:41 <tjones> adding +1 on success, nothing on failure while we triage.  −1 once we are confident
21:08:46 <russellb> even beyond just nova
21:09:21 <russellb> so let us know how things continue to progress
21:09:27 <tjones> will do
21:09:45 <russellb> and I think that's all the project status stuff I really had for this week
21:09:54 <russellb> anyone have any topics for today?
21:10:25 <jog0> russellb: yeah
21:10:41 <mriedem> russellb: yeah, but jog0 go first
21:10:45 <jog0> we should make it clear what is required from a BP detail wise
21:11:08 * mrodden1 lurks
21:11:14 <russellb> yeah, i'm not sure we have any documented guidelines for that ...
21:11:21 <jog0> yet
21:11:24 <russellb> :)
21:11:47 <jog0> one thing  is I think we should have the docImpact details in the blueprint
21:11:59 <russellb> https://wiki.openstack.org/wiki/Blueprints
21:12:09 <russellb> that's the page we should use
21:12:18 <russellb> unless we want to start Nova/Blueprints for some nova specific guidelines
21:12:23 <russellb> and then link over to this one for some general info
21:12:45 <russellb> but I do like the docimpact info on there
21:12:57 <russellb> either on the blueprint itself, or the linked design wiki page
21:12:58 <johnthetubaguy> +1
21:13:07 <jog0> http://justwriteclick.com/2013/09/17/openstack-docimpact-flag-walk-through/
21:13:12 <jog0> is what annegentle wants
21:13:19 <russellb> ok great
21:13:29 <russellb> i think the other thing we need to do is encourage more design info and review up front
21:13:34 <jog0> ++
21:13:44 <russellb> and with some additional people helping review blueprints, it should be more practical to do that
21:14:18 <jog0> ++, thats all for me thanks
21:14:21 <russellb> cool
21:14:26 <russellb> some other things to think about ... prioritization
21:14:33 <russellb> we need to be careful how much we approve at given priorities
21:14:40 <russellb> because we only have so much review bandwidth
21:14:49 <johnthetubaguy> +1 at least set expectations more clearly I guess?
21:15:03 <russellb> yeah, not sure how to best handle it yet
21:15:24 <johnthetubaguy> I liked the medium we track it and talk to people
21:15:32 <johnthetubaguy> low, best of luck if we have time?
21:15:51 <russellb> another problematic case is when people sign up to deliver in icehouse-1, but deliver in icehouse-3
21:15:51 <russellb> and we end up with 80 blueprints in icehouse-3
21:15:51 <russellb> that basically happened in havana
21:15:51 <russellb> and we're *still* catching flak for not merging everything in h3
21:16:09 <russellb> yes, that's the project-wide proposal for how to apply low vs higher now
21:16:14 <johnthetubaguy> yeah, good point
21:16:15 <russellb> but probably need to communicate that well
21:16:29 <johnthetubaguy> I guess we should reject more features down to low, if they slip?
21:16:34 <russellb> +1
21:16:40 <jog0> +1
21:16:44 <russellb> i like that a lot actually
21:16:56 <russellb> you broke the social contract, so it's back to best effort if we can :)
21:16:57 <johnthetubaguy> yeah, agree at the beginning all the medium and above
21:17:12 <johnthetubaguy> then low for everything else from that point, if at all
21:17:18 <leif> How does a group determine which window to put it in?
21:17:30 <russellb> leif: whichever one you think you can have it completed and reviewed for
21:17:30 <johnthetubaguy> when they thing they will get it done?
21:17:34 <johnthetubaguy> think^
21:17:34 <russellb> it's on the developer(s)
21:17:36 <dansmith> everything at medium to start?
21:17:38 <tjones> is there any way to show how complex a BP is likely to be to implement ?  If so, you can push for those to make i-1, i-2 or be dropped?
21:17:56 <russellb> tjones: we can guess :) ... and hopefully the dev proposing it has an idea
21:18:05 <jog0> tjones: hopefully the BP description will say that
21:18:14 <leif> So rules motivate getting a late slot.
21:18:24 <tjones> ok - they should be in the BP desc.  I didn't see a way to explicitly say
21:18:33 <russellb> there's also a downside to getting a late slot
21:18:40 <russellb> much less likely to accept complex/invasive things later
21:18:54 <russellb> and you're also at much higher risk at competing with a higher review load on the backend
21:19:17 <russellb> so it's really in everyone's best interest to deliver as early as is practical
21:19:25 <leif> okay.
21:19:28 <jog0> tjones: there is no specific spot for it, but the BP should cover that level of detail (how it will be implmented risk etc)
21:19:30 <dansmith> so, wait, I got distracted for a sec..
21:19:37 * russellb waits
21:19:45 <dansmith> is the proposal that everything starts at medium and can slip to low if need be?
21:19:58 <dansmith> I thought johnthetubaguy said that, but I definitely don't think that's workable
21:20:00 <russellb> i don't agree everything gets medium automatically
21:20:03 <dansmith> okay
21:20:04 <johnthetubaguy> I think everything we bother to track is medium
21:20:11 <dansmith> right, okay
21:20:15 <russellb> yeah, but still, a lot will be Low
21:20:16 <johnthetubaguy> or higher
21:20:20 <dansmith> gotcha
21:20:26 <johnthetubaguy> yeah, things are low by default I think
21:20:30 <russellb> lots of stuff is "nice to have" but we'd be OK if it didn't make it
21:20:32 <dansmith> I would word that differently
21:20:34 <russellb> ok
21:20:50 <dansmith> not "everything we track is medium or above" because... we're tracking all that stuff in the tracker, even low stuff,
21:21:04 <dansmith> but rather "everything we expect to be mandatory reviews" or something like that
21:21:35 <johnthetubaguy> I guess its important stuff, that got a slot first, that we thing we will have time to review
21:21:36 <dansmith> I mean, we don't have to quibble over words, I was just worried something different was being proposes
21:21:37 <russellb> so ... if it's mandatory reviews ... should we have core reviewers sign up on a blueprint before bumping it above low?
21:21:38 <dansmith> er, proposed
21:21:45 <russellb> should we confirm that we have people willing to "sponsor" it?
21:21:58 <johnthetubaguy> hmm, that would help with the blueprint review, I quite like that
21:22:13 <dansmith> russellb: maybe, that's kinda scary given how many mediums there were
21:22:19 <dansmith> but I can see that for high ones for sure
21:22:20 <russellb> yes, it might mean we have more low
21:22:26 <russellb> but maybe that's a better reflection of reality
21:22:29 <russellb> vs me just making stuff up
21:22:30 <johnthetubaguy> +1
21:22:32 * mrodden assigns all mediums to dansmith by default.
21:22:33 <dansmith> sure, if that's going to make more things low,
21:22:36 <dansmith> then I'm down with that :)
21:22:37 <cyeoh> +1
21:22:50 <johnthetubaguy> yeah, more low things sets the expectation better I think
21:22:59 * dansmith -2's all mrodden's patches
21:23:02 <russellb> and is a way to control how much we actually think we can review
21:23:07 <mrodden> that already happens
21:23:18 <russellb> and *may* even help encourage some companies to put more devs on general work :)
21:23:21 <dansmith> I think this is an excellent thing to shoot for,
21:23:25 <russellb> if they have core folks, they can sign up on reviews, etc
21:23:34 <jog0> I like the idea of having cores assigned to BPs but I think more important is a review of the BP itself
21:23:48 <dansmith> but lets make sure we have a back way out if we start actually reviewing blueprints and decide that it's not going to work to try to put someone's name on everything :)
21:24:21 <russellb> heh, fair enough
21:24:22 <dansmith> so maybe two cores for a medium and four for a high?
21:24:27 <russellb> a nice goal perhaps ...
21:24:38 <russellb> four huh?
21:24:38 <mrodden> ideally, if core's had an incentive to sponsor a BP they would probably have more incentive to review the changes associated
21:24:39 <russellb> bold
21:24:39 <jog0> I think a better BP review (make sure the propsal is good and the design is sound etc) will get us further
21:24:44 <jog0> for less effort
21:25:12 <johnthetubaguy> well, has to pass review by cores to get a priority?
21:25:22 <dansmith> russellb: four to spread out the liability of one person being jammed up and blocking obligatory review of the blueprint because they're half the folks committed
21:25:42 <russellb> could even be 1 for medium, 2 for high, and 3 for essential ... it'd be better than now
21:25:42 <russellb> but i guess 1 doesn't necessarily guarantee it gets the review time
21:25:56 <jog0> dansmith: lets try doing this for I-1 for high or above only and see how it goes
21:25:59 <johnthetubaguy> I like having more than one on all of them, to review the review, etc
21:26:14 <dansmith> johnthetubaguy: exactly
21:26:15 <russellb> jog0: heck, might as well try for the mediums too
21:26:16 <jog0> before we commut fully
21:26:27 <jog0> russellb: sure
21:26:36 <jog0> concerned about bandwidth and flexibility
21:26:40 <dansmith> jog0: well, I think we might as well commit to the whole thing for mediums and above for I1 and just not promise to keep it up until we decide we like it
21:26:41 <russellb> i like that, put this out as an experimental process if icehouse-1 and see how it goes
21:26:49 <dansmith> yeah
21:26:53 <johnthetubaguy> +1
21:27:08 <russellb> awesome!  i like this.
21:27:17 <jog0> works for me.  we should have a BP review party at the summit
21:27:19 <dansmith> I like this right now
21:27:22 * dansmith pulls a comstud
21:27:26 <comstud> lol
21:27:28 <russellb> dansmith: but you reserve the right to hate it tomorrow?
21:27:33 <dansmith> of course
21:27:35 * russellb nods
21:27:38 <dansmith> subject to change without warning
21:27:39 * johnthetubaguy giggles
21:27:42 <jog0> where we divvy up the BPs
21:28:00 <cyeoh> jog0: that sounds good and will encourage people to get BPs in before the summit
21:28:17 <jog0> cyeoh: yeah :(
21:28:21 <dansmith> I don't think we can have them all  filed by summit, many will be created during it
21:28:22 <russellb> skip the drinking and have a nice geek blueprint review party?  :-)
21:28:46 <russellb> many will be created during, and a *bunch*, maybe most, the week after as a result of dicussions and decisions
21:28:49 <jog0> we could start with what we have
21:28:53 <dansmith> we could
21:29:02 <jog0> or just go to the parties
21:29:09 <russellb> i'm usually pretty fried at night, doubt it's that practical
21:29:17 <russellb> i may be hiding in a room to be alone for a few minutes
21:29:35 <russellb> speaking of summit and hanging out ... we should really all try to spend some time hanging out at the summit
21:29:37 <russellb> team building!
21:29:51 <johnthetubaguy1> I would vote for doing something in this meeting the following week? deice what you fancy
21:30:17 <russellb> johnthetubaguy1: yeah, that's fine, but we can dabble a bit in the meantime
21:30:22 <johnthetubaguy1> +1
21:30:28 <russellb> there's not many proposed yet
21:31:06 <russellb> ok, so, lots of improvements for issues we've had wrapped up in this discussion, really great stuff
21:31:07 <russellb> thanks guys
21:31:27 <russellb> mriedem: you had a topic?
21:31:39 <mriedem> russellb: just a call for a core review on a patch,
21:31:46 <mriedem> russellb: https://review.openstack.org/#/c/46718/
21:31:55 <mriedem> that was falling to the bottom of the review pile i think in one of the stats pages
21:31:59 <mriedem> cyeoh looked at it last night
21:32:15 <mriedem> the author has been kind of asking intermittently but his timezone doesn't help much (china)
21:32:15 <russellb> yeah, pretty old it seems
21:32:26 <mriedem> i told him i'd ask here
21:32:29 <dansmith> mriedem: tell him to move
21:32:31 <russellb> nice of you :)
21:32:57 <mriedem> anyway, that's it fro mme
21:33:04 <mriedem> and i got mock working! woohoo
21:33:08 <russellb> heh
21:33:14 <mriedem> you can all enjoy reviewing that soon
21:33:32 <russellb> do you have a sponsor for that patch?  :-p
21:33:39 <mriedem> my mock one?
21:33:48 <russellb> just a joke and reference to the last discussion
21:33:59 <mriedem> oh, missed most of it getting these damn tests to run
21:34:02 <russellb> which was actually only in reference to blueprints, not individual patches
21:34:05 <mriedem> yeah
21:34:12 <russellb> any other topics today?
21:34:13 <mriedem> all i heard was it's ok to dump blueprints in I3
21:34:17 <mriedem> :)
21:34:19 <tjones> LOL
21:34:25 <johnthetubaguy> did we chat about summit sessions already?
21:34:29 <cyeoh> mriedem: and assign then to dansmith
21:34:49 <dansmith> heh
21:34:51 <russellb> johnthetubaguy: not in much detail, just mentioning that today was the deadline
21:34:59 <russellb> johnthetubaguy: did you get my message about meeting tomorrow to start the review?
21:35:15 <johnthetubaguy> russellb: cool, ah, not yet, been out this evening
21:35:40 <russellb> just checked and we have **19** more proposals than time slots (31)
21:35:46 <russellb> johnthetubaguy: hm, was a couple days ago
21:35:46 <russellb> johnthetubaguy: will msg
21:36:40 <russellb> johnthetubaguy: any you want to discuss now?
21:38:47 <russellb> any other topics from anyone?
21:39:27 <mrodden> was there any more news on the midyear meetup?
21:39:33 <russellb> oh, good question
21:39:35 <russellb> ... no
21:39:41 <russellb> and that's my fault, basically
21:39:58 <comstud> dang right it is
21:40:13 <comstud> that's something I can commit to
21:40:13 <cyeoh> we're having a post linux.conf.au (Jan) meetup for those who will be around
21:40:21 <comstud> the statement, not the meeting
21:40:23 <russellb> cyeoh: nice
21:40:51 <russellb> comstud: i can also commit to talking bad about at summit while you're not there to defend yourself
21:41:05 <comstud> :)
21:41:18 <comstud> i'm sure you wouldn't be the first
21:42:48 <russellb> mrodden: so, i need to follow up on that, will work on it
21:42:54 <russellb> thanks for coming everyone!
21:42:58 <mrodden> russellb: k thanks
21:43:32 <russellb> #endmeeting