14:00:57 <johnthetubaguy> #startmeeting nova
14:00:58 <openstack> Meeting started Thu Apr 30 14:00:57 2015 UTC and is due to finish in 60 minutes.  The chair is johnthetubaguy. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:59 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:01:01 <openstack> The meeting name has been set to 'nova'
14:01:02 <andreykurilin> hi!
14:01:04 <jaypipes> o/
14:01:04 <mriedem> hi
14:01:05 <dims> o/
14:01:05 <alex_xu> o/
14:01:06 <dansmith> o/
14:01:08 <bauzas> \o
14:01:17 <johnthetubaguy> welcome all
14:01:21 <ndipanov> o/
14:01:24 <johnthetubaguy> #topic kilo release status
14:01:31 <alaski> 0/
14:01:39 <johnthetubaguy> so this is the last time we talk about kilo, probably
14:01:41 <claudiub> O/
14:02:01 <johnthetubaguy> kilo is out, RC3 has become the final release
14:02:10 <johnthetubaguy> #link https://wiki.openstack.org/wiki/ReleaseNotes/Kilo
14:02:15 <dims> yay
14:02:16 <edleafe> o/
14:02:20 <jaypipes> The Kilo is dead. Long live the Kilo.
14:02:28 <bauzas> \o/
14:02:32 <johnthetubaguy> while its probably too late really, please do check the release notes for correctness
14:02:37 <johnthetubaguy> jaypipes: +1 :)
14:02:53 <johnthetubaguy> we should have had L be lb
14:02:58 <johnthetubaguy> as in pound
14:03:05 <jaypipes> heh
14:03:05 <johnthetubaguy> anyways
14:03:23 <johnthetubaguy> sound you find a bug that takes your breath away
14:03:26 <johnthetubaguy> #link https://bugs.launchpad.net/nova/+bugs?field.tag=kilo-backport-potential
14:03:30 <johnthetubaguy> we have a tag for that
14:03:49 <johnthetubaguy> and hopefully the stable team might jump around and help you through that
14:04:04 <johnthetubaguy> anyways, thanks for all your hard work on kilo
14:04:15 <johnthetubaguy> so…
14:04:24 <johnthetubaguy> #topic liberty release status
14:04:27 <gilliard> \o/
14:04:41 <johnthetubaguy> first things, summit sessions
14:05:01 <johnthetubaguy> we have a nova-drivers meeting and came up with a initial draft of nova summit sessions
14:05:09 <johnthetubaguy> #link https://etherpad.openstack.org/p/liberty-nova-summit-ideas
14:05:31 <johnthetubaguy> now, I guess we need a deadline for session submissions, anyone have a strong preference?
14:05:54 <mriedem> 5/8?
14:06:07 <dansmith> yesterday?
14:06:13 <sdague> yeh, in the past design summit closed out pretty late
14:06:26 <johnthetubaguy> yeah, I was thinking along those lines
14:06:35 <sdague> the schedule was typically set the week before
14:06:45 <johnthetubaguy> so, lets say by the next nova-meeting, so May 7?
14:07:05 <sdague> wfm
14:07:10 <johnthetubaguy> now there is little space left, but you know, we might find something interesting that needs debating
14:07:13 <jaypipes> johnthetubaguy: the sooner we get a schedule up, the sooner we can start moving it around as needed
14:07:17 <bauzas> +1 it would prevent people just coming because session is nice
14:07:28 <bauzas> or having a buzzword in there
14:07:32 <jaypipes> NFV!
14:07:37 <sgordon> dw they will find you
14:07:40 <dims> johnthetubaguy: bauzas: only the fishbowl sessions have to have a title in sched.org
14:07:41 <sgordon> ;p
14:07:45 <johnthetubaguy> jaypipes: feel free to suggest moves on the current one
14:07:58 <jaypipes> johnthetubaguy: ? we have a current shcedule?
14:08:14 <bauzas> jaypipes: ssssht
14:08:15 <johnthetubaguy> jaypipes: I should have been clearer, end of that etherpad, its fully drafted out
14:08:21 <jaypipes> AHHHH
14:08:26 <jaypipes> got it. sorry, missed that.
14:08:47 <johnthetubaguy> #info full draft scheduler for all summit sessions can be found at the end of this etherpad: https://etherpad.openstack.org/p/liberty-nova-summit-ideas
14:08:57 <bauzas> dims: all the nova sessions are fishbowl ones, so...
14:09:22 <jaypipes> bauzas: don't we have working sessions on friaday?
14:09:25 <johnthetubaguy> #info deadline for summit session submissions is 7th May 2015 before the next nova-meeting
14:09:33 <dansmith> jaypipes: it's unstructured
14:09:38 <dansmith> jaypipes: like last time
14:09:40 <sdague> like in paris
14:09:47 <johnthetubaguy> so, we have an etherpad for that too...
14:09:48 <jaypipes> dansmith: right. I think that's what dims was referring to.
14:10:02 <dims> yep
14:10:15 <johnthetubaguy> #info you can find the unstructured meetup ideas for Friday here: https://etherpad.openstack.org/p/liberty-nova-summit-meetup
14:10:27 <johnthetubaguy> I might regret making that public
14:10:32 <johnthetubaguy> so early
14:10:34 <johnthetubaguy> buy hey ho
14:10:39 <johnthetubaguy> but^
14:10:44 <jaypipes> johnthetubaguy: do you have edit rights on that spreadsheet?
14:11:09 <jaypipes> johnthetubaguy: if so, would be nice to color the nova sessions some nice color.
14:11:13 <johnthetubaguy> jaypipes: I have edit delegated edits on sched directly, I can't move rooms
14:11:22 <johnthetubaguy> ah, so apparently thats not possible
14:11:32 <jaypipes> you calling me an arsehole?
14:11:36 <johnthetubaguy> for reasons I don't quite understand
14:11:40 <jaypipes> :P
14:11:47 <johnthetubaguy> nope, I wanted the colors too
14:12:11 <johnthetubaguy> but turns out we need them all the same color, I think ttx knows why, I forget what he told me, sorry!
14:12:28 <johnthetubaguy> OK, cools
14:12:33 <johnthetubaguy> any questions on the summit?
14:12:47 <jaypipes> looks like we'll all be in room 301 for two days straight. bring yoru sleeping bag...
14:12:58 <johnthetubaguy> +1
14:13:00 <johnthetubaguy> so blueprints
14:13:02 <johnthetubaguy> https://blueprints.launchpad.net/nova/liberty
14:13:22 <johnthetubaguy> we are basically not fafing with milestones, so lets keep looking at the above link instead
14:13:24 <johnthetubaguy> #link https://blueprints.launchpad.net/nova/liberty
14:13:34 <johnthetubaguy> anyways, spec reviews are under way now
14:13:45 <johnthetubaguy> lots of stuff already approved and up for review
14:13:50 <bauzas> nit: https://blueprints.launchpad.net/nova/+spec/allocation-ratio-to-resource-tracker has no owner yet, so please add me to it
14:13:53 <bauzas> johnthetubaguy: ^
14:13:59 <jaypipes> johnthetubaguy: the big one for me is online-schema-changes now that jerdfelt is no longer working on nova..
14:14:07 <johnthetubaguy> bauzas: will it not let me
14:14:10 <mriedem> johnthetubaguy: if we aren't doing milestones, what does that mean for feature freeze?
14:14:21 <jaypipes> johnthetubaguy: that really is an important effort. I would be OK having a mirantis engineer pick up that work if cool with others.
14:14:41 <bauzas> mriedem: https://wiki.openstack.org/wiki/Release_Cycle_Management/Liberty_Tracking
14:14:48 <mriedem> jaypipes: oslo.db is mostly mirantis guys already right?
14:14:50 <johnthetubaguy> jaypipes: I am trying to find out about that, lets try sort that out off line, help would be welcome I think
14:15:02 <jaypipes> mriedem: no, not really...
14:15:07 <johnthetubaguy> mriedem: feature freeze is the same, let me get the link
14:15:09 <jaypipes> johnthetubaguy: sounds good.
14:15:10 <mriedem> bauzas: that doesn't say anything about freeze
14:15:33 <mriedem> https://wiki.openstack.org/wiki/Liberty_Release_Schedule
14:15:42 <johnthetubaguy> #link https://wiki.openstack.org/wiki/Liberty_Release_Schedule
14:15:45 <johnthetubaguy> yeah, thats the one
14:15:49 <johnthetubaguy> so the milestones
14:15:52 <johnthetubaguy> we still release them
14:15:57 <johnthetubaguy> we track what get released in them
14:15:58 <jaypipes> johnthetubaguy: k, so what are you asking of us right now on the blueprints? anything in particular you wanted to get decided or done?
14:16:05 <johnthetubaguy> we don't try to predict what will be in them
14:16:14 <bauzas> mriedem: oh, my point was about saying that we're not tracking bps using milestones now IIUC
14:16:27 <johnthetubaguy> jaypipes: really, any questions with the process is where I was going with this, then we have two blueprints to review
14:16:51 <johnthetubaguy> so process, propose your blueprint for the "liberty" series
14:17:01 <johnthetubaguy> that should bring your bp up into that list
14:17:06 <jaypipes> one blueprint I don't see on there that I was hoping to see was the "unfuck-the-neutronv2.py module" one from, IIRC, Brent Eagles.
14:17:07 <johnthetubaguy> if you need a spec, you need to add that
14:17:28 <dansmith> jaypipes: he has a bunch of specs up for that sort of thing
14:17:37 <sdague> jaypipes: I will only review code for that blueprint if that's the actual name :)
14:17:38 <johnthetubaguy> jaypipes: there is a spec for that in review, I think I actually approved that, but its not quite a far ranging in its intitial form
14:17:41 <dansmith> jaypipes: we had a spec last time around that didn't get approved, IIRC
14:17:59 <johnthetubaguy> yeah, its happening, which is good
14:17:59 <jaypipes> so am I just missing the blueprint?
14:18:03 <mriedem> the neutronclient wrapper spec was approved yesterday
14:18:04 <dansmith> johnthetubaguy: right, it's broken into smaller cleanups
14:18:34 <johnthetubaguy> jaypipes: looking
14:19:01 <mriedem> https://review.openstack.org/#/c/131413/
14:19:05 <johnthetubaguy> http://specs.openstack.org/openstack/nova-specs/specs/liberty/approved/wrap-neutronclient.html
14:19:07 <mriedem> johnthetubaguy: jaypipes: ^
14:19:09 <mriedem> https://review.openstack.org/#/c/131413/
14:19:12 <mriedem> wrap neutronclient is easy
14:19:13 <mriedem> https://review.openstack.org/#/c/131413/ is not
14:19:17 <gilliard> https://review.openstack.org/#/c/141129/
14:19:20 <johnthetubaguy> right, thats the other one
14:19:34 <jaypipes> no blueprints for the nova-neutron-refactor...
14:19:51 <mriedem> jaypipes: what is that specifically?
14:20:02 <johnthetubaguy> I think those three are all part of that
14:20:11 <sdague> so.... it seems weird that that spec basically says "the neutron client api is unusable, we'll build you another api"
14:20:15 <jaypipes> mriedem: refactoring of the neutronv2.py module so people can understand it?
14:20:15 <sgordon> he has a couple of specs for refactoring as well
14:20:21 <sgordon> Refactor allocate_for_instance
14:20:25 <mriedem> jaypipes: that's https://review.openstack.org/#/c/131413/
14:20:27 <sgordon> https://review.openstack.org/141129
14:20:33 <sgordon> Refactor of the Neutron network adapter
14:20:38 <sgordon> https://review.openstack.org/131413
14:20:39 <neiljerram> https://review.openstack.org/#/c/141129/ references https://blueprints.launchpad.net/nova/+spec/nova-neutron-refactor, but it appears that that BP link doesn't actually exist.
14:20:41 * beagles arrives late - with ears burning
14:20:41 <jaypipes> mriedem: right. there's no blueprint.
14:20:57 <jaypipes> mriedem: I'm not saying there's no spec. I'm saying there's no blueprint :)
14:21:17 <jaypipes> beagles: I'm very interested in your nova-neutron-refactor spec. wondering where the blueprint might be :)
14:21:18 <mriedem> idk
14:21:31 <johnthetubaguy> jaypipes: yeah, I will try match those up a little, for the merged one, just spotted that
14:21:31 <mriedem> this is kind of a waste of time imo
14:21:34 <dansmith> I usually don't create the blueprint until my spec is close :)
14:21:35 <mriedem> we can find the bp later
14:21:35 <johnthetubaguy> yeah
14:21:43 <johnthetubaguy> lets move on
14:21:48 <jaypipes> whatevs.
14:21:59 <jaypipes> mriedem: I don't see why this is a waste of time :(
14:22:00 <johnthetubaguy> jaypipes: agreed its important for liberty
14:22:07 <mriedem> there are specs
14:22:10 <beagles> jaypipes, quite likely forgot to create the blueprint thing
14:22:10 <mriedem> the bp is process
14:22:28 <mriedem> -1 the spec review for a bp in lp
14:22:29 <mriedem> easy
14:22:39 <mriedem> then we can move on
14:22:41 <jaypipes> :/
14:22:44 <johnthetubaguy> beagles: I guess we are looking forward to the specs being refreshed for liberty for the other neutron refactoring bits
14:23:05 <johnthetubaguy> I think folks are keen to start reviewing them
14:23:07 <johnthetubaguy> which is cool
14:23:08 <bauzas> since it seems there are many BPs, can we track all of them in some unique place ?
14:23:14 <beagles> johnthetubaguy, right
14:23:23 <jaypipes> bauzas: yes, that's called Launchpad. :P
14:23:28 <mriedem> https://review.openstack.org/#/c/131413/ is the big hairy
14:23:30 <mriedem> so that needs review
14:23:33 <johnthetubaguy> bauzas: we track them too many places really
14:23:34 <bauzas> jaypipes: nah, speaking of the network stuff
14:23:49 <johnthetubaguy> lets do that offline
14:23:51 <jaypipes> k
14:24:01 <bauzas> jaypipes: but thanks for the pointer, just discovering it :p
14:24:05 <johnthetubaguy> we have two bps that want to be approved without a spec
14:24:08 <johnthetubaguy> #link https://blueprints.launchpad.net/nova/+spec/convert-image-meta-into-nova-object
14:24:11 <bauzas> anyway, let's move on agreed
14:24:15 <johnthetubaguy> #link https://blueprints.launchpad.net/nova/+spec/network-template-routes-injection
14:24:20 <johnthetubaguy> firstly:
14:24:21 <bauzas> johnthetubaguy: no eventually
14:24:27 <johnthetubaguy> do the need a spec
14:24:35 <jaypipes> not in my opinion.
14:24:39 <johnthetubaguy> secondly, do we like them
14:24:42 <jaypipes> yes.
14:24:46 <jaypipes> I like both.
14:24:51 <johnthetubaguy> are they ready to be approved (i mean)
14:24:56 <jaypipes> yes, IMO
14:24:59 <johnthetubaguy> they seem OK at a quick glance
14:25:00 <ndipanov> first one definitely does not need a spec
14:25:04 <bauzas> johnthetubaguy: https://review.openstack.org/#/c/76234 is finally pointing to an approved spec
14:25:06 <johnthetubaguy> any folks got issues with them?
14:25:07 <jaypipes> so is instance-tagging BP, IMHO :)
14:25:36 <bauzas> but I agree with ndipanov, that's probably even not requiring a spec - or maybe a objects one
14:25:39 <sdague> weren't we trying to get away from file injection for - https://blueprints.launchpad.net/nova/+spec/network-template-routes-injection ?
14:25:55 <jaypipes> sdague: that's not file injection...
14:25:56 <johnthetubaguy> sdague: its using config drive not inject, AFAIK
14:26:03 <sdague> oh, ok, no prob
14:26:31 <johnthetubaguy> anyways, if you have strong feelings please let me know
14:26:51 <johnthetubaguy> I will look to approve them tomorrow ish, assuming there are no big issues
14:27:01 <johnthetubaguy> thanks in advance for you looking
14:27:22 <johnthetubaguy> #help please review the two non-spec BPs ASAP if you are worried about them
14:27:31 <johnthetubaguy> #topic Bugs
14:27:41 <johnthetubaguy> mriedem: hows the gate looking?
14:27:54 <mriedem> i assume good
14:28:03 <mriedem> http://status.openstack.org/elastic-recheck/gate.html
14:28:13 <mriedem> http://status.openstack.org/elastic-recheck/data/uncategorized.html
14:28:16 <sdague> the boot from volume test seems to be failing more often than previously
14:28:17 <mriedem> categorization rate is high
14:28:39 <dansmith> sdague: yeah, noticed that
14:28:58 <mriedem> i opened a bug yesterday for some flaky cinder failures, wasn't boot from volume though
14:29:11 <johnthetubaguy> do we have a ER for that boot from volume at all?
14:29:25 <sdague> so the problem is, it's the generic ssh failure
14:29:27 <mriedem> we have one for the lvm lock
14:29:27 <johnthetubaguy> assuming it was the same reason it failed each time, which might be a rash assumption
14:29:31 <johnthetubaguy> oh…
14:29:36 <johnthetubaguy> gotcha
14:29:36 <mriedem> yeah, the new one is ssh generic failure so we can't really fingerprint it reliabely
14:29:41 <sdague> which we dropped from the signature list, because it's unhelpful
14:29:45 <johnthetubaguy> yeah
14:29:52 <ndipanov> mriedem, lvm lock was that thing in the libvirt driver?
14:29:55 <dansmith> melwitt found something yesterday,
14:30:02 <dansmith> which *could* be related to the ssh thing
14:30:07 <dansmith> ndipanov: no, not libvirt
14:30:08 <sdague> dansmith: do tell
14:30:11 <ndipanov> k
14:30:22 <sdague> because any new breadcrumbs there would be awesome
14:30:25 <mriedem> ndipanov: no
14:30:25 <dansmith> sdague: I thought it was really likely when I first saw it, but now I think it's remote, but it's something
14:30:39 <dansmith> sdague: she realized we're not refreshing info cache from the result of a save() like every other object
14:30:55 <dansmith> and some neutron folks think that sprinkling refresh calls all over the place makes it go away
14:31:03 <sdague> oh, that's interesting. Yeh, seems sensible to fix regardless.
14:31:05 <dansmith> which may just be the timing they're affecting, but this could be a thing
14:31:08 <dansmith> definitely
14:31:17 <sdague> does she have a patch up yet?
14:31:25 <dansmith> dunno, this was late yesterday
14:31:30 <sdague> ok
14:31:33 <dansmith> I'll look in a bit
14:31:38 <bauzas> dansmith: you mean https://review.openstack.org/#/c/178942/ ?
14:31:47 <dansmith> yes
14:31:57 <johnthetubaguy> any more bug related things?
14:31:57 <sdague> once I get through with summit prep next week, I'll try to spend some time digging on some of these bugs
14:32:14 <johnthetubaguy> how is the trivial patch set list going?
14:32:20 <johnthetubaguy> I know its on a kilo etherpad
14:32:23 <bauzas> johnthetubaguy: that's an excellent question
14:32:25 <johnthetubaguy> but we can keeping using that for now
14:32:32 <bauzas> johnthetubaguy: I think we should open another one for Lib
14:32:35 <johnthetubaguy> I noticed it was a bit stale last time I checked
14:32:54 <bauzas> johnthetubaguy: tbh, I think most of us stopped updating that
14:33:00 <dansmith> yeah
14:33:07 <dansmith> burned out quick
14:33:11 <lxsli> the available bugs dropped to a pretty small list
14:33:15 <dansmith> but only because we made real progress I think
14:33:17 <lxsli> and dims was doing such a great job
14:33:40 <bauzas> dansmith: agreed, but kicking-off a new etherpad seems a good signal that this is not just stale
14:33:46 <dansmith> sure
14:33:47 <lxsli> +1
14:34:02 <jaypipes> ++
14:34:05 <gilliard> IDK if etherpad is the best format.
14:34:10 <sdague> yeh, honestly, I expect everyone is going to need a breather and some recharge leading up to summit here
14:34:21 <dansmith> sdague: I was going to suggest,
14:34:21 <lxsli> call it EG nova-trivial-bugs ?
14:34:24 <lxsli> I liked etherpad
14:34:28 <bauzas> gilliard: what are you thinking instead ?
14:34:29 <dansmith> not doing it just yet and letting the bubble reinflate a bit
14:34:38 <bauzas> gilliard: because etherpad worked fine IMHO
14:34:39 <dansmith> so that there is another big chunk of things to attack when we do
14:34:47 <dansmith> because it requires critical mass I think
14:34:48 <sdague> but I agree it would be good to kick off again during the friday meetup
14:34:54 <bauzas> #link https://etherpad.openstack.org/p/kilo-nova-priorities-tracking
14:34:56 <gilliard> spreadsheet eg google docs? We lost a lot of data about what bugs were fixes by that list.
14:35:03 <gilliard> Unless I missed where it went.
14:35:13 <bauzas> could we just stop using gdocs for such tiny things ?
14:35:15 <dansmith> -1 for doing this in a google doc
14:35:20 <johnthetubaguy> #link https://etherpad.openstack.org/p/liberty-nova-priorities-tracking
14:35:20 <bauzas> yeah
14:35:20 <sdague> gilliard: dims was doing rollup counts
14:35:31 <johnthetubaguy> so its there should we choose to use it
14:35:37 <dims> yep
14:35:40 <johnthetubaguy> but giving it a rest for now makes a lot of sense
14:35:43 <gilliard> fair enough.
14:35:56 <dims> need to pick up after say liberty milestone 2
14:36:07 <johnthetubaguy> I was thinking during the friday meetup up
14:36:12 <johnthetubaguy> but lets talk more about it then
14:36:21 <dims> k
14:36:22 <bauzas> +1
14:36:30 <johnthetubaguy> #topic open discussion
14:36:36 <johnthetubaguy> so we have no agenda items here
14:36:40 <dansmith> so I have a thing
14:36:41 <johnthetubaguy> are we all done?
14:36:47 <johnthetubaguy> dansmith: fire away
14:36:54 <dansmith> This is now good with t-h: https://review.openstack.org/#/c/174480/
14:37:01 <dansmith> recommend we slam in with extreme prejudice
14:37:23 <mriedem> done
14:37:25 <johnthetubaguy> dansmith: yes, that probably makes sense
14:37:28 <johnthetubaguy> cool
14:37:48 <jaypipes> ++
14:37:51 <bauzas> ++
14:37:51 * mriedem wants to get out of 2 meetings at once
14:38:12 <johnthetubaguy> so all done?
14:38:28 <johnthetubaguy> thanks all, talk again soon
14:38:36 <johnthetubaguy> #endmeeting